Tuatara: Volume 14, Issue 3, December 1966
Vegetative features of Griselinia lucida— — A New Zealand shrub Epiphyte
Vegetative features of Griselinia lucida—
A New Zealand shrub Epiphyte
Introduction
Large, Dark-Green shining leaves, usually contrasting strongly with those of the host trees, and distinctively fluted roots descending to the ground make Griselinia lucida or ‘Puka’ the most conspicuous shrub epiphyte in the New Zealand rain forest.
G. lucida is distributed throughout the North and South Islands, but is much more common in the north. The other New Zealand species of the genus—Griselinia littoralis—has a similar distribution, with the addition of Stewart Island, but is more common in the south and at higher altitudes. The two species also contrast in growth habit. G. lucida appears to be primarily adapted to an epiphytic role, but in common with most vascular epiphytes, even the most specialised tropical forms, it can also occur terrestrially on open rocky sites. G. littoralis is usually terrestrial, but in Stewart Island and elsewhere it is sometimes epiphytic.
Griselinia has a special plant geographical interest as it belongs to a small group of genera which exhibit a wide disjunction in their distribution between South America and New Zealand. There are five species of Griselinia in Chile and some of them are reported to be shrub epiphytes. The family placing of the genus is uncertain but it is currently included in the Cornaceae.
Foliage
The leathery mature leaves of Griselinia lucida are usually large and can be classified as ‘mesophylls’ in Raunkaier's terminology. In this respect they agree with the majority of tropical rain forest species. The leaves are often as wide as long and are usually markedly asymmetrical about the midrib, the distal portion of each leaf relative to the branch bearing it being both longer and wider than the proximal portion. In very young leaves the laminae are symmetrical about the midrib, so the asymmetry must result from unequal growth as the leaf develops.
page 122
Fig. 1: Griselinia lucida on kahikatea (Podocarpus dacrydioides). Root descending to the left, foliage at base of host crown. Above the foliage of the Griselinia is a mass of Collospermum hasfatum. Plateau Reserve, Hutt Valley.
Agfa Isopan Ultra. Dev. I.D. 11. 8½ min. All photographs taken with a Linhof camera (4 × 5″) with 150 mm. f. 5.6 Schneider Symmar Lens.
The undersurface of the lamina contrasts with the shining upper surface by its whitish opaque appearance. This is due to densely concentrated stomata which extend uniformly over the undersurface except at the position of the midrib. The thick cuticle covering the lower epidermis arches over each stoma to form a distinct cavity. Beneath the upper epidermis, which also has a thick cuticle, there are several layers of isodiametric cells lacking chloroplasts which Oliver (1930) suggests may function for water storage. Most of these leaf features can be regarded as modifications serving to reduce water loss, such modifications being common among epiphytes which are as a group much more subject to drought than plants establishing on the forest floor.
Most vegetative growth ceases during the the winter and the immature leaves at branch tips are protected by usually two green and fleshy bud scales. The latter are equivalent to the sheaths of ordinary leaves together with rudimentary petioles and laminae no more than a millimeter or two long. The scars left by these bud scales can be seen in Figure 3.
Another unusual feature also shown in Figure 3 is the siting of lateral shoots at some distance above the subtending leaves. Again dissection of vegetative buds shows that the lateral shoots are situated in leaf axils at an earlier stage, but are apparently carried upward by subsequent elongation of the internodes.
Epiphytic Growth Habit
Griselinia lucida can be found as an epiphyte on a wide range of host species including the tall emergent conifers — Dacrydium cupressinum (rimu) and Podocarpus dacrydioides (kahikatea) — and smaller main canopy species such as Elaeocarpus dentatus (hinau), Beilschmiedia tawa (tawa), Dysoxylum spectabile (kohe-kohe) and Vitex lucens (puriri). It seems to establish only on mature trees in the forest, presumably because only mature trees have branches sufficiently well illuminated for the epiphyte to survive.
Fig. 2: Roots of two Griselinia lucida shrubs on kohekohe (Dysoxylum spectabile) hosts. Waikanae.
Agfa Isopan Record. Dev. Rodinal (Dil. 1-100) 20 min.
The young roots are closely appressed to the host bark, frequently growing into crevices and behind bark flakes. The root tips are white, smooth and opaque, but from a short distance behind them the root surfaces are densely clothed with short root hairs. These apparently persist until cork formation commences as they are undiminished even several feet from the growing points. Where the roots are in contact with the host bark they are anchored by the root hairs and the union is sometimes so complete that when the roots are pulled away they either remove portions of bark or leave strips of their own tissue behind. The function of the root hairs on the exposed parts of the roots is more puzzling. If they remain capable of water absorption, however remote from the root tip, then the young plant would be able to take full advantage of every rain shower. The root hairs are so closely set that they might also function in a sponge-like manner to absorb and retain water for some time and also act to reduce water loss by evaporation in much the same way as does the tomentum of a leaf.
page 126In a particular case where the roots of a small plant had not reached the ground they were found to have extended for a distance of 4 feet in the course of a year. At this rate of growth it would take several years for the roots to reach the ground from the branches of some of the taller host trees.
Fig. 4: Young roots of Griselinia lucida on rimu (Dacrydium cupressinum). Kaitoke, near Wellington.
I ford F.P.3. Dev. I.D. 11. 7 min.
The bark of the mature roots is striking because the longitudinal grooves are usually parallel and continuous. Anatomical investigation shows that the groove pattern is directly derived from that of the primary vascular rays. This can be made clear by a consideration of the development of the tissues of the root. In the young root there are usually about 8-14 primary xylem strands surrounding a wide pith which later becomes lignified. When the vascular cambium becomes active, as well as forming secondary xylem with large scattered vessels and secondary phloem with alternating layers of fibres and thin walled tissues, it also forms wide primary rays leading from the primary xylem strands, which become a permanent and conspicous feature of the root cross section. The first cork cambium arises in the pericycle, as is usual in roots, but subsequent cork cambia arise in each parenchymatous secondary phloem layer after it ceases to function for conduction. The tissues outside the innermost cork cambium are dead and comprise the bark. The cork cells derived from the cork cambia are thin walled so most of the rigidity of the bark is due to the persistent fibre layers of the secondary phloem. Indeed the bark differs very little in its appearance from the secondary phloem. The longitudinal grooves of the mature root arise from splitting of the bark in the positions of the primary rays following on increase in diameter of the vascular tissues. Further increase in root diameter results in widening of the splits into grooves V-shaped in cross section.
The formation of cork results in loss of the root hairs, among other tissues, and the enlarged roots usually become separated from the host trunk, hanging free much in the manner of liane stems. The high single main root, however, frequently gives rise to a special type of lateral root, already mentioned, which grows horizontally completely around the host trunk. Such ‘girdling’ roots may be produced in great numbers and branch freely to form a tangled network about the host (Fig. 5), but usually many of these die away leaving only a few to enlarge by secondary thickening. Their function would appear to be to bind the epiphyte firmly to the host.
Fig. 5: Griselinia lucida on kohekohe (Dysoxylum spectabile). Note the girdling roots. Waikanae.
Agfa Isopan Record. Dev. Rodinal (Dil. 1-100) 20 min.
My own observations would suggest that although Griselinia lucida usually sends roots to the ground it probably never becomes a self-supporting tree on death of the host, but falls with it. Oliver (1930) states however that Griselinia lucida ‘generally grows as an epiphyte high up on forest trees, but the larger specimens send roots to the ground and thus establishing themselves grow into independent trees’.
Apart from the occasional adoption of the epiphytic habit by Griselinia littoralis, G. lucida appears to be the only shrub epiphyte on trees other than tree ferns in New Zealand, which sends roots to the ground, but does not become self-supporting. In this regard it is interesting to note that in New Caledonia there also seems to be only one shrub epiphyte with this habit — Fagraea berteriana, and in North Queensland only Brassaia actinophylla. Thus in respect of this particular type of epiphyte the New Zealand rain forest is on an equal footing with the rain forests of New Caledonia and North Queensland.
References
Oliver, W. R. B., 1930. New Zealand Epiphytes. Jour. Ecol. 18: 1-50. Raunkiaer, C., 1934. The Life Forms of Plants and Statistical Plant Geography. Clarendon Press, Oxford.