Go to National Library of New Zealand Te Puna Mātauranga o Aotearoa
Volume 56, 1926
This text is also available in PDF
(2 MB) Opens in new window
– 423 –

Forest-floor Covering and its Life.

[Read before Wellington Philosophical Society, 27th August, 1924; received by Editor, 5th December, 1924; issued separately, 10th April, 1926.]

Nature and Variations of the Covering.

In almost all forests, “bushes,” or even shrubberies—at any rate, in New Zealand—there is present above the surface of the solid ground a layer, more or less thick, and varying greatly in texture and composition, of flowers, leaves, twigs, and other debris undergoing decomposition. It is to this layer, grading upwards, in a typical case, from “humus” through “mould” to partly-disintegrated and finally newly-fallen and quite-undecayed fragments, that the term “forest-floor covering” is here applied.

At first sight it might appear to be just so much litter of no practical importance, which in a cultivated forest might be raked up and burned. This, however, would be quite an erroneous idea, for its functions in the economy of the natural forest are various and far-reaching. Before considering these, however, a glance may be taken at the covering itself. It varies greatly from place to place in the same forest. Here it is a foot or more in thickness, there but an inch or less; in one place it is covered with “filmy ferns” and mosses, in another bare, loose, and dry; while its constituent fragments may vary from logs to tiny leaf-fragments. Yet it has two constant characteristics: it is all resting passively on the ground, and it is all undergoing disintegration and decay.

Different types of forest have different floor-coverings. This is due largely to the nature of the debris falling on to the floor, but differences of situation, climate, and moisture all induce corresponding differences in the floor-covering. That of a swamp forest will often be thick and peaty, while that of a hillside forest will often be thin, loose, and well rotted beneath. The age of the forest also makes a difference, old forests usually having thicker mould than young ones. Many other factors doubtless contribute to cause the differences observed—such as the chemical and mechanical composition of the soil. An example of this has recently been recorded in Nature by J. B. Farmer, who found that in certain English woodlands real leaf-mould of any thickness formed only where the surface soil was of a sandy or gravelly character; whereas where it was of a heavy, especially calcareous, nature the leaves simply rotted down and evaporated, so to speak, by the next year. He attributes this to the increased activity of bacteria in the presence of calcium carbonate causing decomposition, mainly to carbon dioxide and water. In this case the determining factor seems to be the degree of acidity of the soil. It is also possible that the increased supply of lime salts in the one case greatly aided the activities of the earthworms, to the proper digestive processes of which animals this element in fair amount is a necessity. Thus not only may the bacteria have been able to work upon the debris more rapidly, but this process may have been aided and abetted by the earthworms.

According to Colville (1), true leaf-mould is alkaline, due to the residual lime (carbonate) after the organic acids have been washed out, evaporated, or otherwise dissipated. When one year's leaf-fall is not converted into

– 424 –

leaf-mould (alkaline) by the following year, the process may be suspended by the leaching into it of acids from the freshly fallen leaves, and the removal of lime salts by the roots of plants, and a condition of permanent acidity result. The product in this case is not true leaf-mould, but a substance called “upland peat” because of its resemblance to bog peat. It is defined as “a nonpaludose deposit of organic matter, chiefly leaves, in a condition of suspended and imperfect decomposition and still showing its original leaf-structure, the suspension of decomposition being due to the development and maintenance of an acid condition which is inimical to the growth of the micro-organisms of decay.”

Functions of the Covering.

Turning to the functions of the debris covering, we may consider them from two aspects, the chemical and the physical.


As plants grow they abstract from the soil certain elements which are present normally in rather small amount. These are distributed to all parts of the plant, and, although some of them are partly reabsorbed into the body of the plant before the leaves fall, a good deal is retained by these organs when they die and are shed. Also, when a branch or a whole tree dies, material is locked up inside in a form that cannot be directly utilized by the higher plants. Hence in time the soil would become depleted were no process of disintegration and decay of debris, with liberation of plant-foods, occurring. It is mostly in the debris stratum on the forest-floor that this transition takes place. Many organisms, as we shall see, take part in it, besides some perhaps purely chemical changes, such as direct oxidation. At the bottom of the layer much of the material is in a form suitable for utilization by plants, and fresh material is always working down as the old becomes decayed, oxidized, and absorbed in the soil.

Colville (1) remarks: “In soils poor in lime, trees and other plants constituting the vegetative mantle of the earth may be regarded as machines for concentrating lime at the surface of the ground. The lime is drawn up by the roots in dilute solution from lower depths, is concentrated in the foliage, and the concentrate is transferred to the ground by the fall and decomposition of the leaves.”

There is another aspect, somewhat more obscure. It is known that bacteria are of enormous importance to the fertility of the soil, especially in relation to what is known as the “nitrogen cycle.” Some fix free nitrogen from the air, others render nitrogen available in the soil from otherwise unavailable sources, while these and others may under other conditions use up and so diminish the supply of nitrogen available (2). On the whole, however, their activities are beneficial, and at present, at any rate, are certainly indispensable. Now, to do this work they require a source of energy, which is most easily obtained from carbohydrate material. This, in the forest, is almost all supplied by the floor-debris. A point of interest in New Zealand forests is that many of the timber-trees bear on their roots very numerous nodules of the size of small shot and smaller, and these may be found in enormous numbers throughout much of the more decayed portion of the mould. Though their nature is disputed, * whether they

[Footnote] * See article by J. S. Yeates in july, 1924, number of Journal of Science and Technology.

– 425 –

be comparable to the nodules on the roots of legumes which house special nitrogen-fixing bacteria, or whether they be purely of fungal origin, it is suggestive that they occur in such numbers where the supply of carbohydrate is great. It may be that this is essential to their well-being, and therefore almost certainly to that of the trees.


The humus supplied by the decay of this debris is of great value in bringing about a good physical texture of the soil. Every one knows this to be necessary for cultivated soils, and it is no less true for forest soils. It lightens the soil, making the penetration of roots easier. It also makes the soil warmer and more equable; more porous, allowing greater penetration of surface water; and more absorptive, thus causing greater retention of moisture. In this way the bad effects of drought are ameliorated and floods are greatly reduced.

The net result is that soils of all types of original texture, from heavy clay to sandy and gravelly soils, tend to become more uniform in texture and better adapted for a timber covering. For purposes of regeneration the soil is now better adapted than in the original condition, for seedlings grow much more easily in it than on hard soil.

Animals of the Forest-floor.

Having considered the forest-floor covering itself, it may now be considered from the point of view of its suitability and use as a dwelling-place, shelter, ambuscade, and source of food for the individuals of a large and varied animal community. That this community exists is easily verified. One has only to disturb the leaf-mould or turn over a log, and “hoppers,” centipedes, spiders, insects, and all manner of creatures scatter in every direction. Their invisibility under ordinary conditions is partly due to the form and colour of those which live on the surface resembling their surroundings, but in the case of the great majority is due to their manner of life, for these are the Cryptozoa—animals which spend their existence burrowing in dark crannies and tunnels of the under-strata of the mould and in the soil itself, completely hidden from cursory view, but nevertheless carrying on great and innumerable activities. It is the nature of these activities and the animals responsible for them that are now to be considered.

Method of Study.

There are two ways of seeking information on this subject. The one may be called the qualitative or field-naturalist method. This consists in making excursions to as many types of forest as possible at all times and seasons, turning over logs, mould, and stones, and noting the species of animals occurring, their frequency, and distribution. This is an excellent method, and has been adopted by Shelford (3) in the United States with great success, and by many naturalists in New Zealand. It suffers from several serious defects, however. In the first place, few can pretend to such a detailed knowledge of all the many forms of animal-life occurring in this situation as to attempt to identify by sight more than those belonging to one or two orders at the most; and, even if specimens are collected to be identified by specific authorities, of all those appearing new and unusual to the investigator many are missed for lack of knowledge sufficient to

– 426 –

pick out the small details by which an expert separates many of the forms so similar in general appearance. Again, it is difficult to gain a true idea of the relative numerical importance of the various elements in the fauna, as one is inclined to give more notice to those that are most aggressive in their activity, or are largest, or most evenly distributed. Many small or protectively-coloured species are liable to be overlooked, and there are various minor disadvantages.

The other method, which was mainly adopted by the writer, is from time to time to take representative samples of the forest-floor, carefully sort and look over them, and pick out and preserve all the animals occurring. In this way a truer idea can be obtained of the identity and distribution of the animals, and various numerical and seasonal relationships shown. However, there are disabilities in this method also. The great local variations in the floor-covering, the difficulty of knowing to what depth to go, and the presence of large fragments like logs make if difficult to get a really representative sample. Then, as bright sunlight must be used when picking over the sample to stir otherwise invisible animals into activity, and a large amount of time is consumed in the operation, samples cannot be taken as frequently as is desirable. Moreover, almost nothing is learnt of the habits and life-histories of the members of the animal community; that must be gained by careful and patient observation in the forest itself.

While, therefore, relying mainly on the latter method, the writer endeavoured to apply the former as much as possible.

The floors of two types of forest only were studied in detail. These were respectively mixed rain forest, represented by Chapman's Bush (adjoining Wilton's Bush, Wadestown), and southern-beech forest, represented by that on the hill above Days Bay. An area of about 50 yards square was marked out in each locality, and representative samples taken from these once a month from March to September, inclusive, 1923. Ten small samples, each 9 in. square, were taken monthly from each locality and combined to make two large samples. These were sorted over and all the large animals, together with sticks, &c., removed, the remainder thoroughly mixed and divided evenly till only one-eighth was left. This was then very carefully sorted over till all animal-life visible in the brightest sunshine was removed. No doubt a certain number of the more minute were inevitably left behind, but those found were surprisingly numerous. The number picked out from the small sample multiplied by eight, plus the number picked out at first, gave a rough estimate of the number originally present. All the specimens were identified as nearly as possible with the ready aid of various authorities and the results tabulated. Though too long to reproduce these here, a summary is given at the end of this article of the species mainly occurring. Only the commoner speciesr generally speaking, were identified with anything like certainty. This was partly due to the fact that the literature on many of the groups is some what scattered and difficult to obtain, while many of the more obscure species of the less-studied groups are not yet described.

Relation of Forest-floor Animals to their Environment.

In nature the great majority of animals may be said to fall into one or other of three classes, according to the way in which they are adapted to their environment. The first comprises those which show their colours openly, making no attempt at concealment. These are the possessors of some special power or qualities which give them racial protection: such as

– 427 –

a high rate of reproduction; some quality rendering them distasteful or inaccessible to most other animals—for example, the hairy caterpillar of the magpie moth; great size and strength; or the possession of armour.

In the second class are those animals which, by dissimulation of some kind or other, seek to escape observation either for purposes of offence or defence. The great majority of animals living exposed to the light come under this heading.

The third class neither publish their existence nor disguise themselves, but live in hidden stations away from the light. These, as before mentioned, are sometimes called Cryptozoa. Having external protection in the form, say, of some part of a plant, a decaying log, the soil, o even, in the case of parasites, of another animal, and being consequently more or less in darkness, they have no use either for protective colouring or form. As a result colour is more or less dispensed with, and the form which approximates to that of the greatest utility in such situation is adopted. Generally speaking, this form is cylindrical, with legs and other appendages reduced to the minimum size consistent with use (4). This is well illustrated in the case of the earthworm and of the burrowing larvae of certain insects—for example, the wood-borers.

These classes are not, of course, hard-and-fast. Many animals are on the border-line between one and another, or may change from one to another at different stages in their life-histories. But in broad outline they indicate definite tendencies in animal-life, and in many cases there is no doubt as to where we should place a given animal.

Thus the great majority of the inhabitants of the forest-floor debris are dull-coloured creatures with few peculiarities of shape or structure and of a distinctly retiring disposition. Most of them shun bright light—they run from it on every occasion; in other words, they are negatively phototactic. Most of them like a considerable degree of moisture, and gradually retreat farther from the surface during dry weather. In these two respects, however, there is a great degree of difference observable between the various members of the forest-floor community. Thus Parorchestia sylvicola (the ordinary bush-hopper), most of the earthworms, and the long slender centipedes which feed largely upon them (species of Zelanophilus), and in general the pale translucent-looking animals, are very sensitive to light; while many of the spiders, mites, beetles, and in general dark or well-marked and opaque-looking ones like the large woolly ground-spiders Hexathele hochstetteri and Porrhothele antipodiana, do not fear bright light nearly so much, and in consequence live near or on the outer surface of the floor-debris. Similarly, those with hard (chitinized) integuments usually require less moisture than the thinner-skinned species, and live frequently in drier situations. Earthworms, the larvae of some insects, hoppers, woodlice (Isopods), and many others, can only thrive where there is a considerable degree of moisture.

Although all living together in the same medium, the food habits of the animals under consideration present almost all possible variations. These may, however, be summarized under three headings. There are those which live on the medium itself—in other words, on decaying vegetable and animal remains (saprophagous animals); those which eat their fellow denizens, or make predatory raids into the strata above (arboreal) or below (subterranean) that in which they live (carnivorous animals); and those which feed upon roots, leaves, or other parts of plants (phytophagous). There are no sharp boundaries between these classes, as many of the

– 428 –

animals are omnivorous, or change from one class to another at different stages in their life-histories. The majority, however, fall either exclusively or predominantly into one or other of the classes.

Saprophagous Animals.

As regards the saprophagous animals, in the first place the fresher fragments of the debris are eaten by many of these animals. Much indigestible material is often thus ingested for the sake of the small digestible portion it contains. The latter, consisting of certain protein, carbohydrate, and some of the mineral matter, is abstracted, and after use by the animal ultimately returned to the soil in a condition readily available for plants. The woody undigested residue passing from the animal as faeces is in a finely pulverized condition, very suitable as a medium for fungal and bacterial action, by which means it is ultimately broken down into its simpler constituents. The earthworms, daddy-long-legs larvae (Tipulids), and “hoppers” (Parorchestia sylvicola) illustrate this process. Again, by constant boring and burrowing operations in search of food, many animals pulverize the undecayed fragments as well as keeping them open to the air and well turned over, thus aiding bacterial action. Finally, in the case of earthworms especially, a great deal of the surface debris is carried fairly deep below ground in burrows, and earth is brought to the surface, thus causing a gradual mixing of the whole, with a hastening of the process of decay and an enrichment of the lower layers of the soil. In this way less nitrogen is lost than would otherwise be the case (5).

The following quotation from Colville (1), whose article has come to my notice only since the above was written, is interesting in this connection:—

“The chief agents in the decay of leaves are undoubtedly fungi and bacteria. There are other agencies, however that contribute greatly to the rapidity of decay. Important among these are earthworms, larvae of flies and beetles, and myriapods, or thousand-legged worms. Animals of all these groups exist in myriads in the leaf-litter. They eat the leaves, grind them, partially decompose them in the process of digestion, and restore them again to the soil, well prepared for the further decomposing-action of the microscopic organisms of decay.

“The importance of earthworms in hastening the decay of vegetal matter was pointed out long ago by Darwin. The importance of myriapods, however, as contributing to the formation of leaf-mould, has not been adequately recognized. In the cañon of the Potomac River, above Washington, on the steeper forested talus slopes, especially those facing northward, the formation of alkaline leaf-mould is in active progress. The purer deposits are found in pockets among the rocks, where the leaf-mould is not in contact with the mineral soil and does not become mixed with it. The slope directly opposite Plummer's Island is a good example of such localities. Here during all the warm months the fallen leaves of the mixed hardwood forest are occupied by an army of myriapods, the largest and most abundant being a species known as Spirobolus marginatus. The adults are about 3 in. in length and ¼ in. in diameter. They remain underneath the leaves in the daytime and emerge in great numbers at night. On one occasion a thousand were picked up, by Mr. H. S. Barber, on an area 10 ft. by 100 ft., without disturbing the leaves. On another occasion an area 10 ft. by 20 ft. yielded 320 of these myriapods, the leaf-litter in this case being carefully searched. Everywhere are evidences of the activity of these animals in the deposits of ground-up leaves and

– 429 –

rotten wood. Careful measurements of the work of the animals in captivity show that the excrement of the adults amounts to about ½ c.c. each per day. It is estimated on the basis of the moist weight of the material that these animals are contributing each year to the formation of leaf-mould at the rate of more than 2 tons per acre.”

Other Forms of Life.

Although they cannot be considered in detail here, brief mention must be made of other forms of life present in the debris stratum. Among these are the microscopic unicellular animals, the Protozoa, whose activities are mainly concerned with limiting the number of bacteria; and the much larger composite animals belonging to the same phylum or natural division, the Mycetozoa, sometimes called “slime-fungi”—curious thin, jelly-like plates of living matter which creep slowly over damp rotting material, feeding both saprophytically upon it and also digesting any small organisms in their course (2). Then, besides ferns, mosses, and algae, which, being green plants, do not much concern us in this connection, there are present a large number and variety of fungi. These feed saprophytically and help to a great extent in breaking down the debris, as well as serving as food for much of the animal-life.

It is well that all the agencies that assist in bringing about the dissolution of the fallen trees of the forest cannot operate in the artificial environment of our wooden dwelling. The main factor controlling most of them is perhaps the absence of moisture. In the forest there is a natural sequence, commencing probably with the wood-eating beetles and other insects (huhu grubs, for instance), which, having tunnelled the dead trunk in all directions and reduced much of the material to powder, leave the field clear for tipulid larvae, millipedes, amphipods, and others; and these again, having served their turn, give place to the final scavengers, the earthworms. It must not be thought, however, that this is a simple linear process of one animal living on another's faeces, for, throughout, bacteria and fungi are continually transforming these remains, as well as building part of them into their own bodies, the latter in turn being used as food by subsequent denizens.

Life Strata in the Forest.

The debris-dwelling animals have so far been considered as a separate community. It remains to consider them in relation to the rest of the forest inhabitants and the life of the forest itself.

The animals of the forest dwell in three main strata, each exhibiting certain characteristics. These strata are—(1) The subterranean stratum; (2) the forest-floor-debris stratum; (3) the arboreal stratum. In each of these strata live animals exhibiting the three main types of feeding-habit enumerated previously, but the relative abundance of each kind of food in each of the strata determines to a great extent the relative proportions of these animals present. Thus in the soil itself roots and fine scattered food-particles left in the decay of plant-remains form the principal source of food; in the debris stratum dead material and the life it supports (abundance of saprophagous animals) are the principal supply; while in the arboreal stratum food is abundant in the form of foliage and other plant-tissues, and this is reflected in the large number of plant-feeding animals, which, again, support a predatory population.

– 430 –

None of the communities of these three strata is, however, independent. Each is necessary to the others, and depends, in part immediately and in part ultimately, upon them for its existence. Thus in the subterranean stratum the rather fixed and well-distributed food-material is mostly eaten by slow-moving burrowing animals, which usually simply eat their way through the soil, having little need of special-sense perceptions to guide them, leaving to their digestive apparatus the task of separating the assimilable portions. The obstacles in the way of carnivorous animals gaining a living in this stratum are, however, considerable. They must either be endowed with a special sense enabling them to track their victims, or be contented to turn over a relatively enormous amount of material in a random search. Consequently we find carnivorous animals in the minority.

In the forest-floor stratum conditions are very different. The most abundant food-supply is the litter of decaying leaves, stems, and other vegetable material, which is generally fairly loose; it thus supports not only a numerous saprophagous population, but also affords good concealment and opportunities of ambush and chase for a large and varied predatory one. Living vegetable material is here at a minimum, consequently also the animals which feed upon it. This stratum, which lies in a position of vantage midway between the other two, and offers abundant opportunities for concealment, is the habitat of a large number of the carnivorous animals of the forest. These obtain their food not only from the stratum in which they live, but also at the expense of casual immigrants into their stratum—such, for example, may be the fall of insect-larvae from trees, their descent into the ground to pupate, the migration of nocturnal animals for concealment by day, or the emerging of insect adults from the subterranean stratum. Some, such as many opiliones and spiders, augment their food-supply by incursions (chiefly nocturnal) into the arboreal stratum. Thus they probably exercise a considerable degree of control over the phytophagous species, which are largely harmful. This is perhaps especially the case with wood-boring insects. Arboreal carnivorous species (excepting the parasitic ichneumon wasps) mainly confine their attention to external plant-feeders; but, in accordance with their habit of living in confined surroundings, such carnivorous units as centipedes, spiders, and opiliones of the forest-floor stratum may be found in tunnels and crevices in living wood, and the last have actually been observed by the writer eating wood-boring insect-larvae.

Quite a different aspect of their role in the forest life is their use as food by a large number of the forest-dwelling birds. Their number being great, and not so dependent on season or weather as in the case of the arboreal population, they form an important winter food reserve, though many of them are less palatable than the leaf-eating grubs and the ripe berries obtainable at other seasons.

Finally, it may be said that, as they form an integral and natural part of the general life-community of a forest, it is clear that both in the above and in many more obscure ways they must contribute greatly to that healthy position of equilibrium which obtains in its maturity.

Description of Forests and Localities.

Beech Forest.

The first of the two localities studied is a portion of forest, approximately 50 yards square, lying on the south-west aspect of the beech-forest-covered hills on the east side of Port Nicholson, near the top of the ridge behind

– 431 –

the Days Bay pavilion. The altitude is about 1,000 ft. The forest is the original southern-beech association of these hills, and is composed of the following trees:—

Forest-trees: Nothofagus Solanderi, N. fusca (both very plentiful), Weinmannia racemosa (very common), Metrosideros robusta, M. florida (liane), Elaeocarpus dentatus, Podocarpus ferrugineus, Griselinia littoralis (small tree), Suttonia salicina (small tree), S. Urvillei, Coprosma grandifolia (small tree), Knightia excelsa, Drimys axillaris (small tree).

Undershrubs: Coprosma foetidissima, C. Colensoi, C. lucida, C. rhamnoides, Cyathodes acerosa, Leucopogon fasciculaius, Dracophyllum Urvillianum var. filifolium, Nothopanax Colensoi, Cyathea dealbata.

Floor: Polypodium grammitidis, Cyclophorus serpens, P. Billardieri, Hymenophyllum demissum (with several other species, very common), Trichomanes reniforme (common in patches), Blechnum capense, Lindsaya viridis, Asplenium flaccidum, Dianella intermedia, Astelia nervosa, Pterostylis Banksii, Gahnia spp., Uncinia filiformis and other species.

The forest-floor is covered in many places by a dense growth of filmy fern (Hymenophyllum spp.) and kidney-fern (Trichomanes reniforme), while mosses, liverworts, and fungi are present in patches. Undershrubs are present but are not closely spaced. Above the ground-level is in most cases 1 in. to 3 in. of decayed vegetable matter, held together with a network of fine roots, while above this is a fine layer of loose debris (leaves, twigs, &c.). The subsoil, which comes close to the surface, is mostly a greywacke rock.

The situation is on a fairly steep slope, and the dense foliage of the beeches combined with the southerly aspect renders it somewhat darker at the ground-level than is the case in the second locality. Owing to the above factors and its altitude (1,000 ft.) it is also considerably colder. The exact conditions are given in connection with each sample taken.

Tawa Forest.

The second locality is a similarly-sized portion of forest in the middle of the remnant of a formerly more extensive rain forest to the south-west of Wadestown, known as Chapman's Bush, now a portion of the Wellington Corporation reserve. This is largely a tawa forest. A list of the plants occurring in this limited area as given by Mr. B. C. Aston is as follows:—

Forest: Beilschmiedia tawa, Elaeocarpus dentatus (hinau), Olea montana, Metrosideros robusta (rata), Knightia excelsa (rewarewa).

Small trees and shrubs: Pennantia corymbosa (kaikomaka), Metrosideros florida (liane), M. scandens, Griselinia lucida (epiphyte), Rhipogonum scandens (liane), Parsonsia heterophylla (liane), Astelia Solandri (epiphyte), Suttonia australis, Coprosma grandifolia, C. robusta, C. crassifolia, C. areolata, C. rhamnoides, Macropiper excelsum, Geniostoma ligustrifolium, Brachyglottis repanda, Cyathea dealbata, Myrtus bullata.

Floor: Dryopteris velutina, Blechnum filiforme (epiphyte), Asplenium lucidum, A. flaccidum (epiphyte), A. adiantoides (epiphyte), Polypodium Billardieri.

The material of the forest-floor covering is generally much looser, less decayed, and less matted by roots than is the case in the first one. The altitude is only about 100 ft. above sea-level. Owing to less density in the

– 432 –

tree-foliage, and its situation on a gentle slope facing north, it is also considerably lighter and warmer than the first locality. Ferns are not so numerous on the forest-floor, but there are paiches of the climbing-fern Blechnum filiforme, and several others. Mosses, lichens, and fungi are also comparatively scarce, and in many cases dry debris alone forms the upper surface of the floor.

Though surrounded by forest, this area is less removed from settlement than the former one.

Date of taking Samples, and Weather Conditions.

(A = beech forest; B = tawa forest. The figures in the second column refer to the numerical order in which the samples were taken.)
Mar. 17. 1A. Ground damp from rain some days previously; day bright, with clouds.
Mar. 23. 1B. Ground damp, rain in morning; day cloudy.
April 13. 2B. Ground dry, no rain for some days; sky overcast.
April 20. 2A. Ground very dry, no rain for over a week; day bright and hot.
May 11. 3B. Ground wet after week's rain; day dull; some rain, warm.
May 18. 3A. Ground wet, rain during morning, sunny at noon when sample taken; warm.
June 13. 4B. Ground very wet from rain during week previous and in morning; sunny at noon, warm.
June 20. 4A. Ground wet, but weather fine till midday, when cold rain commenced. Sprouting noticeable in seeds and roots of sample.
July 13. 5B. Ground damp; weather cold but fine.
July 20. 5A. Ground very wet; day cold and showery.
Aug. 13. 6B. Ground damp; day sunny, warm.
Aug. 17. 6A. Ground fairly dry; day fine and warm.
Sept. 7. 7B. Ground wet, rain in morning; day warm, and sunny.
Sept. 14. 7A. Ground fairly dry; day warm and humid.

Fauna of the Forest-floor Covering.


1. Turbellaria.—Land planarians (or flatworms) do not appear to be represented to nearly the extent that might be expected. Many of this group, however, are not at all or only partially cryptozoic, and the majority appear to be rather restricted as to habitat. Thus some species may be found in damp portions of wood, or under the bark of decaying logs, which, owing to the difficulty of including them in a representative sample, and also to their being frequently the loci of quite local concentrations of the forest fauna, were perforce neglected in the present study. Those planariáns found, however, are fairly distributed among the samples from both localities.

Geoplana sanguinea, the species occurring in the greatest number, was found three times in samples from each locality. It is one of the dullest and least diversified in colour of our native planarians, and in this respect agrees with the more typical Cryptozoa, with which I think it may be included.

The other species found were G. subquadrangulata, G. spectabilis, and G. agricola. They are perhaps more in the nature of casual invaders. All are carnivorous.

– 433 –

2. Oligochaeta (Earthworms).—These occurred in considerable numbers in all the samples. In general they are more numerous in the beech-forest series than in the rain forest. Their numbers fluctuate rather widely, conditioned probably to a large extent by the state of the debris as regards moisture content, which in turn is greatly influenced by the immediate weather conditions. It is well known that earthworms come nearer to the surface in wet weather: thus in the forest occurs a temporary invasion from the more subterranea pnpoulation into that of the surface stratum. In the September sample from Chapman's Bush locality occurs a great increase, perhaps in a large measure of this nature. The bushes were actually dripping water when the sample was taken. There seems to be no variation of a progressive kind indicated throughout the period.

The soil in Chapman's Bush is much deeper and more loamy than that of Days Bay ridge, a fact which is probably correlated with the larger number and, it may be stated, the much larger size of the earthworms present.

Species of Maoridrilus and of Octochaetus were found in both localities; Diporochaeta spp. occurred only in the Days Bay samples and Neodrilus spp. only in the Chapman's Bush ones. A considerable number of minute worms could not be identified.

The members of this group are typically saprophagous.

3. Crustacea.—The Crustacea are among the most completely cryptozoic and most typical members of this type of association. This is especially true of the amphipod Parorchestia sylvicola, which appears to be distributed in the floor-debris of almost all types of forest and scrub associations in New Zealand. It is also distributed rather evenly, and its numbers do not appear from field observations to fluctuate much throughout the year. The adults were not found to fluctuate much in number in either locality throughout the period. Their widest variations are between 20 and 66, which is not so great considering the many arbitrary factors involved. The young show, as might be expected, greater variation.

There appears to be indicated a rather regular monthly alternation between high and low numbers of the young. Possibly this is connected with some brood-production period, but the figures are insufficient to generalize upon.

The males of this species are known to be rather rare. All the specimens collected were therefore determined for sex, the large flattened front legs of the male being, on the suggestion of Dr. Chilton, used as a guide. It was found that in a total of 383 specimens from the beech-forest locality, 43, or 11.2 per cent., were males, while in a total of 333 from the tawa forest no males were discovered, though each specimen was carefully examined with a dissecting-lens. Moreover, the males in the former were fairly evenly distributed throughout all the samples. It would thus appear that males are extremely rare or are not produced in the lower-altitude forest during this portion of the year. Temperature, dependent on altitude, is suggested as the environment factor most likely to be involved.

This species is typically saprophagous.

Isopoda: Among isopods (slaters, or woodlice) much greater differences in distribution are apparent. They are very much more abundant in the Chapman's Bush samples than in those from Days Bay, the totals being respectively 491 and 86 for the period, while the species present seem each to be restricted to one or the other locality. Trichoniscus spp. occur in both localities, while Philoscia spp. (pubescens) occur only in the Chapman's Bush samples. Species identified were T. thomsoni, T. phormianus, P. pubescens, and a species of Cubaris.

– 434 –

There is noticeable in the Chapman's Bush figures a considerable increase in the number of isopods present during the winter months of June, July, and August, with a slight decrease in September. The maximum occurs in July. This is more noticeable in the actual specimens, which are mostly mature and of large size during these months, while they are much smaller in the preceding months. It is probably the dampness of the debris, due to frequent rains and little evaporation, that is the chief factor in the case.

The isopods are both saprophagous and phytophagous.

4. Chilopoda (Centipedes).—The chief representatives of these typical Cryptozoa were Anopsobius neozelanicus and Paralamyctes validus (scarce) among Lithobiomorpha, and various species of Zelanophilus and Zelanion among the Geophilomorpha. The first of these two types was almost equally common in both localities; the second is considerably more common in the beech-forest samples, the individuals also being larger. There appears to be a tendency to increase in numbers and in size of the Chilopoda during the winter months. The Scolopendromorpha are decidedly rare, being represented by only two individuals collected.

The members of this group are all carnivorous.

5. Diplopoda.—There being no systematic work on the New Zealand Diplopoda, these typical saprophagous and phytophagous Cryptozoa are classified in the tables only under families. Of these, various species—Julidae, Polydesmidae, and Siphonphoridae—occurred rather commonly The last appear to be represented more commonly in the beech-forest samples (by a species of Siphonophora); the others show little difference in this respect. No progressive changes in the population appear to be indicated.

All the above groups are whole-life-cycle Cryptozoa typical of this type of association in general. With the exception of certain members of each, this cannot be said of the succeeding groups.

6. Insecta.—Among Insecta a very miscellaneous collection is represented.

Collembola (“springtails”) are noticeable by their small numbers, mostly occurring in Chapman's Bush samples.

Thysanura were not found. This is decidedly surprising when compared with the debris-fauna of European pine forests as investigated by Pillai (6), who states that this group and Nematodes are by far the largest ones represented. From field observations the writer can add that this appears to apply as regards Collembola to pine forests in New Zealand.

It is worthy of note in passing that no Nematodes (thread-or round-worms) large enough to be detected with the unaided eye were found in any of the samples. Some water infusions of portions of the material, however, revealed a few microscopic ones.

Orthoptera: With the exception of some small undescribed forficulids (“earwigs”) and an odd cockroach, few were found. The rarity of wetas is noticeable; but the same consideration applies here as in the case of planarians. Identified were Cutilia sidilotti, Onosandrus pallitarsis, and Argosarchus horridus.

Lepidoptera (moths) and Diptera (two-winged flies), where found, are almost certainly casual immigrants. Only an odd specimen or so occurred, including a small Simulian and a Cecidomyiid.

Hemiptera (bugs): Certain of these are typical inhabitants of leaf-mould, notably Metagerra obscura (mainly in beech forest), which is widely

– 435 –

distributed. Their numbers are not relatively large. Coelostomidia zealandica, Targarema stali, Hahnia australis, a Myodochid nymph, and specimens of a new genus of Cicadellidae also occurred.

Hymenoptera: In this order occur the most anomalous differences found in the investigation. Ants, being colonial, are not amenable to any ordinary method of sampling. Prenolepis longicornis is the species most commonly found, and in some samples constituted an overwhelming majority of the fauna. It appears to be more common in samples from Days Bay than in those from Chapman's Bush. Other ants were Ponera castanea (rare), Aphaenogaster antarcticus, and a species of Formica.

A small ichneumon found is perhaps parasitic on one of the insect larvae of the association.

Coleoptera (beetles): Those that are not perhaps casual invaders are probably the Staphylinidae (including Quedius sp.), Pselaphidae (Sagola, Euplectus, Gastrobothrus spp., and others), and Carabidae (Agonochila binotata, Oopterus, and Metaglymma sp., and others unidentified) among carnivorous forms; Bitoma, Enarsus, and Pseudopatrum spp. among the saprophagous ones; and weevils among the phytophagous ones.

Sagola occurs only, and rather commonly, in the Days Bay samples; various “carabs” also follow the same rule. The staphylinids are more common in Chapman's Bush samples; weevils are much more common in the former than in the latter.

Also identified were—(Elateridae) Corymbites sp.; (Dascillidae) Veronatus sp.; (Byrrhidae) Pedilophorus sp.; (Lamellicornia) Odontria sp., Ontho-phagus posticus; (Hydrophilidae) Tormissus magnulus, (Cucujidae) Dryocora howittii; (Heteromera) Hylobia nubeculosa, Pseudopatrum tuberculicostatum; (Longicornes) Somatidia antarctica; (Rhynchophora) Phrynixus astutus, Phemus scabralis, Nonnotus sp., and Brachylous sp.

Insect larvae (grubs or caterpillars) constitute a large proportion of the population. Most of the insects concerned are Cryptozoa only at this stage of their life-history. Among the carnivorous ones carabid or staphylinid larvae are present in both localities throughout. Elaterid larvae are also common. Saprophagous forms are found mainly among the Diptera. They occur commonly throughout both series. A large brown tipulid larva is particularly evident in the April sample from the tawa forest.

One species of Lepidoptera larvae is particularly common throughout the Days Bay samples, occurring to a lesser extent in the Chapman's Bush samples. It is probably a root-feeder, being always found well distributed throughout the matrix of fine roots. A great number of larvae could not be placed with any certainty.

7. Arachnida.—The Araneida (spiders) are represented by certain characteristic members, while others are probably only chance visitants. Of the former, Hexathele hochstetteri and Lycosa umbrata are common, the former more so during the first four months. Both are evenly distributed between the two localities, the latter species being the commoner. A large species of the family Dysderidae occurred regularly in the Days Bay samples, doubtfully in the other ones. Clubiona peculiaris, a very pale-cream-coloured spider, was scarce but characteristic of Chapman's Bush locality. Sidyma angulata, perhaps a surface-dweller, occurred at times in both. Araneus crassus occurred once, and a Salticid twice. Many of the small spiders are probably immature Hexatheles and Porrhotheles (antipodiana), though several seem to be adult micro-araneids. The members of this

– 436 –

group are present in only moderate numbers, but by their carnivorous habits and frequent large size exercise a considerable influence.

Opiliones (harvest-spiders) have two chief forms present—a species of Purcellia represented by moderate numbers, and several closely-allied species of Triaenonyx, most common in the rain forest. Other cryptozoic species are known to occur in both areas, but are probably more specialized, not venturing far from certain habitats. They are also mainly nocturnal in their movements. Identified were (Laniatores) Triaenonyx coriacea, T. cockayni, and (Palpatores) a species of Macropsalis.

Pseudoscorpions (book or false scorpions), represented by several unnamed species, are also more common in the rain forest.

Both these groups are carnivorous.

The Acarida (mites), in point of numbers, form a very large proportion of the associations of both localities. This is probably also much larger than is indicated by the tables. Considering their minute size, however, their relative importance is considerably reduced. Many species are present, but as none, or practically none, are named, any indications of genera shown are purely tentative. Beetle-mites are by far the most numerous, some being of relatively large size. The species of the Rhyncholophidae shown in the rain-forest samples is a very pretty mite of a bright-red colour, about ⅛ in. diameter. It has four or five pairs of long, stout, latero-dorsal finger-like processes, extending from the level of the first pair of legs to the posterior extremity, beset with very numerous long hairs, many of them pectinated. These intertwine, giving the animal a woolly appearance.

Most of this group are probably either parasitic or predatory. The following occurred: Oribatidae—Oribatella sp. (common), Nothrus sp. (common, especially in beech forest); Haplodermidae—several spp., Gaima-soidea (several species, especially in beech forest); Troglyphidae—several species; Eupodidae—only in tawa forest (rare); and Rhyncholophidae (rare).

8. Mollusca.—The Mollusca (snails) are represented by several species, chiefly of the genera Endodonta and Laoma. All are small, and none can be called common. Their food is probably derived from the low-growing herbage. Identified (tentative) were E. biconcava, E. tapirina, E. coma; E. pseudoleidon, E. anguiculus, Laoma marina, and Lagochilus lignarius.


There is nowhere any great divergence between the fauna of the two localities; all the main units are represented, and, leaving out of account the ants, to about the same extent in each locality. When we come to genera and species, however, certain of these appear to be restricted to each habitat, and, among those that are common to both, rather wide variations in numerical importance become apparent. This is in accord with expectation: the two habitats are similar in their main environmental conditions; in both cases it is the floor of a forest of considerable age practically in its primeval condition; there is no great geographical separation; and the remaining factors, except that of the plant covering, show no great variation. With the plant covering it is largely a case of substitution of one ecological unit for an equivalent or almost equivalent one in relation to the remainder of the association, but adapted suitably to the slight differences in the rest of the environment. Somewhat similarly with the animal association, substitution of one unit for a similar one occurs without affecting appreciably the character of the association as a whole.

Considerable differences occur in the composition of the faunas of each locality, as indicated by the samples taken from month to month: this

– 437 –

is probably due to several causes—in part to a real seasonal change, and in part to particular weather conditions. Thus the large increase in the number of earthworms in some samples appears to be largely due to wet weather immediately preceding causing them to come nearer the surface, and so be included in the sample.

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

Monthly Totals of Larger Taxonomic Groups.
(A = beech forest; B = tawa forest.)
Month. Insects.
Coleoptera. Ants and Ichneumons. All other Insects.
Carnivorous. Saprophagous. Phytophagous.
A. B. A. B. A. B. A. B. A. B.
March 51 40 32 10 80 16 204 281 53 3
April 144 64 26 16 32 126 64 49 20
May 82 32 1 20 16 36 9
June 176 40 8 16 1,304 88 8 25
July 320 41 34 1 24 8 20 20 43
August 88 16 1 41 40 1 2,040 64 35 25
September 216 57 72 1 48 68 11
Total, all months 1,077 233 151 76 264 26 3,762 581 212 125

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

Month. Insect Larvae. Acarida. Opiliones and Pseudo-scorpionida. Araneida.
Carnivorous. Saprophagous. Phytophagous.
A. B. A. B. A. B. A. B. A. B. A. B.
March 62 34 97 104 17 264 176 11 75 34 65
April 44 12 50 32 42 6 160 481 50 88 113 59
May 20 3 3 11 14 73 42 10 33 78 32
June 35 43 70 21 10 136 169 12 67 67 47
July 88 8 35 60 53 12 640 249 44 61 36 17
August 64 44 90 80 34 18 376 96 40 173 95 78
September 48 1 24 74 29 7 272 184 32 91 66 20
Total, all months 361 145 296 423 207 67 1,921 1,397 199 588 489 318

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

Month. Chilopoda. Diplopoda. Crustacea. Mollusca. Oligochaeta. Turbellaria.
A. B. A. B. A. B. A. B. A. B. A. B.
March 54 58 71 53 192 191 10 10 42 53 1 2
April 126 131 71 10 71 97 8 32 28 56
May 123 40 81 12 155 87 32 8 21 54 2
June 117 222 21 98 111 263 3 41 79 113 2 2
July 193 89 46 66 186 249 32 107 85 54 1
August 201 157 30 90 164 244 51 58 70 115
September 204 65 108 12 108 183 48 49 36 356 1
Total, all months 1,018 762 428 341 987 1,314 184 305 361 801 6 5
– 438 –

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

Monthly Totals of Larger Ecological Groups (excluding Ants).
(A = beech forest; B = tawa forest.)
Month. Carnivorous. Saprophagous. Phytophagous.
A. B. A. B. A. B.
March 477 450 434 411 160 29
April 637 835 246 211 131 58
May 388 182 258 156 79 31
June 545 590 211 552 48 76
July 1,321 466 386 430 129 170
August 864 564 355 570 160 102
September 839 361 333 625 160 57
Total, all months 5,071 3,448 2,223 2,955 867 523

Some General Considerations.

Although the present subject is really an ecological one, no attempt has been made above to deal with it in strictly ecological terms. The debris-dwellers of two types of forest were studied, and for a limited time only, in detail. Other types in various localities and at various altitudes would almost certainly reveal quite different animal associations, and these might possibly be classified on a broader basis into formations. Besides the indigenous forests, those of exotic origin certainly have very modified associations. This is certainly true of Pinus radiata forests, but what the quantitative relations are has not been tested. A huge field for investigation is open here, especially as many of the animals are so imperfectly known.

In the large general debris stratum of a forest the animals do not all bear an equally close relationship. Many small spheres of intimately connected activity are present. The term used by Shelford (3) for such a direct interdependence is a “consocies.” The interdependence may be the result of reliance on the same specific food-supply, or it may be more in the nature of symbiant associations. Thus we might speak of the earthworm consocies as being constituted by the earthworms and the various animals, such as flat-worms (Planaria sp.) and centipedes (Zelanophilus sp.), which feed more or less particularly upon them. Again, we have the consocies constituted by the ant Prenolepis longicornis and the various species of Hemiptera (bugs) which are associated more or less symbiotically with it.

With regard to statistical results, the complete statement in tabular form of all the occurrences of the animals found is too lengthy for publication, but an analysis of these has been attempted and the results are given above under the headings “Monthly Totals of Larger Taxonomic Groups” and “Monthly Totals of Larger Ecological Groups.”

The uncertain position of the Formicidae, which form a large proportion of some of the totals, prevents any close comparisons. Where present in excessive numbers they may also have reduced the numbers of other animals by the time the latter could be separated. The numbers of those classified as “carnivorous” are considerably augmented by the mites, which are small and only partly carnivorous. Their irregularity in numbers is also responsible for the larger fluctuations observable in this group.

Taking into account these two considerations, it will be seen that the numbers of animals present throughout the period does not alter to any

– 439 –

considerable extent; that carnivorous and saprophagous units constitute by far the greater part of the associations, and are present to about the same extent; and that phytophagous units are of relatively small number.

Comparison with Similar Communities in other Coutries.

It is difficult to gain any definite evidence regarding the forest-floor populations of other countries. Shelford (3) gives some general descriptions of several such communities in temperate North America, but in rather a casual manner, and with scarcely any mention as to relative importance of the various groups. From what one can gather, however, the differences between the floor-fauna of these (deciduous) forests and of New Zealand ones must be considerable. Amphipods, the one species of which is perhaps the most characteristic animal of the latter, are absent; isopods receive comparatively little mention; molluscs (snails and slugs) are comparatively quite common; spiders are numerous, and certain centipedes and millipedes are mentioned, but seemingly insects are by far the most numerous inhabitants.

A more exact research has been made on the floor-fauna of pine-forests in Bavaria by Escherich and Pillai (6), an abstract only of which is available to the writer at present. The following summary is given:—

“The results of these patient investigations were collected in tables, and show that a great number of small organisms, some of them characteristic and others accidental, are to be found in the plant debris carpeting the soil of pine-woods. Among the first are nematodes, aptera, and acarians—these occur in enormous quantities, there being often many thousands per square metre; Cecidomyia larvae—several hundreds per square metre, and then in considerably fewer numbers (5–100 per square metre); Arachnida, Annelida, Chilopoda, Formicae, Diptera, &c.”

The results are strikingly different from those obtained in the present investigations, but have been discussed in connection with Aptera above.

It must be remembered that the results given in connection with the present investigation apply to only seven months of the year, and that results obtained from a summer study of the same localities might be considerably different. This is to be anticipated, seeing that the ground would on the average be considerably drier and warmer.

Though scarcely comparable, some researches of Morris (7) on the fauna of arable land and of pasture at Rothamsted are of interest in this connection. On arable land it was found that half the fauna consisted of various insects; next most abundant were earthworms, Collembola, ants, millipedes, mites, and centipedes.

From a research on a forest-floor and on meadow-land near Washington, U.S.A., McAtee calculated that the number of animals per acre belonging to Arthropoda, Annelida, and Gasteropoda for forest was 1,216,880, and for meadow 13,654,710. A rough calculation of the numbers present in the two localities studied here shows that there are about 13,700,000 in the beech forest, and about 8,700,000 in the tawa forest. These are given for comparison, though, according to Buckle (8), “except where there is a pronounced infestation, the estimate of numbers per acre has purely a fictitious value.” The writer is inclined to think, however, that though this may be the case when applied to single species, it is scarcely so when applied to an approximate computation of the whole fauna.

In conclusion, the writer would gratefully acknowledge assistance received from the following gentlemen who kindly aided him in the identification of

– 440 –

the animals concerned: Professor W. B. Benham, Mr. G. Archey, Mr. G. V. Hudson, Mr. T. Cockroft, Mr. J. G. Myers, Mr. W. R. B. Oliver, and especially Professor H. B. Kirk for his interest, suggestions, and criticism. To Mr. B. C. Aston the writer is indebted for naming the plants, and for general interest and criticism. Finally, he would record his indebtedness to his wife for her enthusiastic help throughout, especially in the laborious task of picking out and preserving the small animals.


1. Frederick V. Colville. The Formation of Leaf-mould. Jour. Washington Acad. Sci., vol. 3, No. 3, Feb., 1913.

2. Sir E. J. Russell, F.R.S., and Others. The Micro-organisms of the Soil. London. 1923.

3. Victor E. Shelford, Ph.D. Animal Communities in Temperate America. 1913.

4. Herbert Mace, F.E.S. The Evolution of the Caterpillar. Science Progress, April, 1923.

5. Charles Darwin. The Formation of Vegetable Mould through the Action of Worms. London. 1881.

6. K. Escherich; Dr. Pillai. The Animals living in the Plant Debris covering Forest Soil. Review in International Review of the Science and Practice of Agriculture (n.s., vol. 1, No. 1, p. 85, 1923) of Die Streufauna, by K. Escherich in Forstwissenschaffiches Centralblatt, pt. 1 (Berlin, 1922), and of work by Dr. Pillai.

7. Hubert M. Morris, M.Sc. On a Method of separating Insects and other Arthropods from Soil. Bull. Ent. Res., vol. 13, p. 197, 1922.

—– Observations on the Insect Fauna of Permanent Pasture in Cheshire, Annals of Applied Biology, vol. 7, Nos. 2, 3. Dec., 1920. Insect and other Invertebrate Fauna of Arable Land at Rothamsted, ibid., vol. 9, Nos. 3, 4, Nov., 1922.

8. Philip Buckle. On the Ecology of Soil-insects on Agricultural Land. Jour. Ecology, vol. 11. No. 1, May, 1923.