Other formats

    TEI XML file   ePub eBook file  


    mail icontwitter iconBlogspot iconrss icon

Tuatara: Volume 20, Issue 2, March 1973

A Guide to Morphological Characteristics for use in the Description of Extant and Fossil Dicotyledonous Leaves

page 75

A Guide to Morphological Characteristics for use in the Description of Extant and Fossil Dicotyledonous Leaves


A plea is made for more adequate descriptions of extant leaves as an aid for overseas and New Zealand palaeo-botanists who have no or limited access to New Zealand material. As a guide, a list of preservable dicotyledonous characters is given. An experiment on the identification of extant leaves carried out in 1968 is briefly discussed in an appendix.

It is confirmed that the ratio of entire to non-entire leaf margins forms an important ecological indicator that is not invalidated by a lack to taxonomic determinations.

Generally speaking the taxonomic descriptions of plant taxa deal inadequately with the morphology of their leaves. Venation characteristics, which include some of the main features used in the identification of plants from leaves, are often omitted. For example, the description of leaves of Griselinia littoralis reads 'lvs broad-ovate to ovate-oblong, rounded at apex, less glossy, slightly to not unequal-sided at base; main veins hardly prominent to obscure below' (Allan, 1961, p.443). The generic description tells us that the leaves are alternate, very coriaceous, entire and exstipulate with petioles dilated and slightly sheathing at the base (p.442). This is a rather full description compared with that given for some other species. But what is the venation like? What is the size of the leaf? Can this description be used to separate this leaf form from similar leaves? I think not.

I do not wish to press this point too hard because arguments can be put forward for better descriptions of other parts of plants, e.g. seeds, fruit, wood anatomy and morphology. There are, however, two main reasons why, from a palaeobotanical viewpoint, I consider it necessary to improve leaf descriptions of extant plants: —


The abundant fossil angiosperm leaves, in Quaternary and Tertiary sequences, are very difficult to study without detailed knowledge of the morphology of extant leaves. Reference material and literature are not always available.


Study of fossil leaves may lead to a better understanding of the present ecological distribution of species when used as a supplementary technique to palynology in investigations of past distribution and palaeoclimatology.

One of the most accurate ways of describing leaf characters is by illustration. Clearly it would be impractical to include photographs and/or drawings of all species in the various books on the page 76 New Zealand flora, but it would be of great help if as wide as possible a range of variation in leaf form could be illustrated. It is often more advisable to describe a particular feature by illustration than to describe it in writing. This is particularly so in the descriptions of leaves, both fossil and extant, where terminology is so variable that different taxonomists use different terms to describe the same feature. In many cases descriptions and illustrations of fossil material have been so inadequate (although still acceptable within the framework of the International Botanical Code) that they have been disregarded by subsequent authors who have not had access to the original material.

Problems in Leaf Systematics

This section is essentially a review of an excellent paper by J. A. Mouton (1966/1967) which is not apparently well-known to New Zealand botanists, supplemented by work done by myself in 1968. Madler and Straus (1971) have done similar work and have prepared a list of 109 characters for use in the description of leaves.

The study of leaf systematics poses many complicated problems that are not discussed here in any great detail; this has already been done elsewhere by Mouton (1966/1967) and Wolfe (1971).

Firstly, there are problems inherent in both fossil and extant leaf material:

  • 1. Variability in size and shape:

    This may be caused by growth deficiencies or excesses, periods of arrested growth, polymorphism and heterophylly. The New Zealand flora has a high percentage of juvenile leaf forms, some forms lasting well into the adult stage of tree or shrub growth. Some 10 per cent or about 200 species have this character (Millener, 1960). Also, species of many New Zealand genera are unusually diverse, both morphologically and ecologically, e.g. Coprosma, Hebe, Senecio, Plagianthus, etc.

  • 2. Convergence of forms from different families:

    Parallel development may be inherited or environmentally induced. Many of the strange features of the New Zealand flora involving parallelism have been listed by Millener (1960). They include cushion plants, climbing plants, epiphytes, divaricating plants and others, often occurring in genera which do not exhibit such habits elsewhere. The conifer-like appearance of the whip-cord hebes is a further example of convergence.

    Secondly, there are problems peculiar to fossil leaves:

  • 3. Loss of characters by fossilisation:

    This problem could be rather serious in New Zealand since tectonic activity, even in Quaternary rocks, has caused small scale page 77 movements within sediments, sometimes resulting in the enclosed fossils becoming contorted and split beyond recognition.

  • 4. Distinguishing compound leaves from simple leaves.

  • 5. Present method of identification:

Fossil leaves are simply compared with extant herbarium material without detailed study of either.

Many characters used in the description of fossil leaves can also be used to identify extant leaves. These are as follows:


Well-preserved fundamental characters:


Primary and secondary venation




number of pairs of secondary veins


distance between veins


mode of attachment of secondary veins to midrib


angle of emergence of secondary veins (where parallel)


density (see Manze, 1968).


Tertiary venation




angle of emergence from base of secondary veins.


density (see Manze, 1968).


General leaf form and position of midrib.


Primary and secondary domatia.


Leaf margin.


Well-preserved characters of secondary importance:


Leaf biometry


surface area


length to breadth ratio.


Morphology of base and apex.


Secondary intercalary veins.


Geometry of the areole and the intersecondary field.


Poorly-preserved characters:




length of leaf blade


apical ratio (length of drip-tip to length of leaf blade)


Basal ratio (length of base to widest part of leaf blade to the length of leaf blade).


Petiole length and presence or absence of a pulvinus.


Internal glands and cystoliths.


Hairs and stomata.


Phyllotaxy and stipules.

These characters can all be recorded in a punch card system. A preliminary attempt has been made to do this for all extant New Zealand dicotyledonous species. At this stage the cards only contain characters used in published descriptions so that little is yet known about venation details. Only a few species are based on herbarium specimens as the N.Z. Geological Survey does not page 78 have an adequate herbarium and study of herbaria elsewhere has not yet been possible. Even at this stage interesting information can be obtained relatively easily from such a system although it is not yet possible to identify all ‘unknown’ native leaf forms by this system.

There are about 1,300 extant dicotyledonous species in New Zealand, including the off-shore and Sub-Antarctic islands. Disregarding altitude and overlapping ranges (both geographically and in the size range of the leaves) the distribution of species and their margins and leaf size classes can be set out as follows:
MarginLeaf Size ClassesMarginMarginMarginMarginMargin
Approximate Total No. of SpeciesNot entireSometimes entireAlways entireLeptophyllNanophyllMicrophyllMesophyllMacrophyllMegaphyll
North Island6652930412646432431
South Island995312346314537182-
Stewart Island265283042325248212-
Chatham Islands1023025451944552631
Sub-Antarctic Islands893728352150401881

Further, and probably more meaningful, results can be obtained by dividing the lowland trees into their botanical districts and working out the distribution of various leaf classes from north to south.

The percentage of plants with particular characteristics or combinations of characteristics can be readily worked out. Of particular interest to palaeobotanists and palaeoecologists are the percentages of plants with leaf morphology associated with particular climatological and ecological niches. For example, plants in tropical areas generally have large (mesophyllous) leaves with long drip-tips, irrespective of their position within the various forest strata. In a normal tropical rain forest about 70-80% of the species fall into this category (Richards, 1952; Wolfe, 1971). Taking the overall flora in New Zealand and off-shore islands there is a decrease in the percentage of species exhibiting acute apices from the Kermadecs to the Sub-Antarctic Islands. The following table shows that the percentage does not reach more than 60% suggesting that the New Zealand forests could not be regarded as tropical in this respect. The percentages could reflect tropical affinities as they are characteristic of sub-tropical floras, especially those listed by Wolfe (1971).

page 79
Table listing percentages of species with acute apices. The percentage of species that exhibit acuminate and/or acicular apices are shown in brackets (based on information in Allan, 1961).
LocalityPercentage of acutely apexed species
Kermadec Islands60 (29)
North Island52 (15)
South Island49 (13)
Stewart Island48 (15)
Chatham Islands46 (14)
Sub-Antarctic Islands37 (14)

If the same and similar studies could be applied to fossil leaf material then comparisons could be made with the present day climate. In this case it would not be necessary to identify the leaves but just to distinguish between different leaf forms. A preliminary attempt has been made on a fossil flora from Whakarongo Stream, which drains into the Waikato River (N51/f894). No serious attempt was made to identify the leaves but twenty-five different leaf forms were recognised.

The marginal conditions and leaf size ranges are given in the following table and are compared with present day data given by Dawson and Sneddon (1969) for a forest covering the Maungataniwha Ranges, near Kaitaia, some 280 km to the north.
Leaf SizesMargins
Fossil flora8.5%37.5%50%4%75%25%
Kaitaia Forest6%68%25%1%60%40%

The percentages of the marginal characteristics of the Kaitaia Forest include compound leaf types which were not distinguished in the fossil flora. The fossil flora has more mesophyllous and entire leaves than the Kaitaia flora, suggesting warmer temperatures, and the temperature difference would presumably be even greater at the locality of the fossil flora.

For further assessment of the palaeoclimate, the pollen flora was studied to see if it supported the climatic interpretation based on the leaves. The pollen also assisted in determining the age of the rocks. The pollen spectrum was dominated by extinct beeches (Nothofagus) of the 'brassi' group (some of the fossil leaves were of extince ‘brassi’ group beeches with typically broad leaves).

page 80

In fact 47% of the total pollen flora was of this pollen form, suggesting warmer temperatures than at present in the same locality, since the Nothofagus ‘brassi’ group is now found in New Guinea and New Caledonia in warm moist conditions (although in climatic zones no different from North Auckland today). The presence of this group, as well as other extinct species — Haloragacidites harrisii (Couper) Harris, H. trioratus Couper, Proteacidites minimus Couper (cf. Knightia excelsa R. Br., which was found as a leaf fossil), Tricolpites alveolatus Couper and Polypodiidites inangahuensis Couper — suggested a Pllocene age (probably Opoitian or maybe Waitotaran). The full pollen flora is listed as follows:

Slide number L5736. Locality N51 / F895

Adopted floral age: Opoitian to Waitotaran
Species ListCount
Cyathea spp.21
Monolete spores24
Histiopteris incisa1
Dicksonia squarrosa2
Paesia scaberula1
Trilete spores2
Pyrrosia serpens2
Phymatodes diversifolium4
1Polypodiidites inangahuensis Couper2
Dacrydium cupressinum2
Phyllocladus sp.2
Podocarpus sp.1
*Nothofagus matauraensis Couper100
*N. cranwellae Couper8
?*N. spinosus Couper1
*N. waipawaensis Couper1 derived Teurian fossil
N. ‘fusca’ group4
*Proteacidites sp.1
*P. minimus Couper (cf. Knightia excelsa)1
*Haloragacidites harrisii (Couper) Harris8
*H. trioratus Couper1
Myriophyllum sp.3
Epilobium sp.3
Fuchsia sp. (cf. F. excorticata)1
Myrtaceae2 (leaves found)page 81
unidentified tricolpate and tricolporate species4
*Tricolpites alveolatus Couper1
Pomaderris sp.2
?Nertera sp.1
Coprosma sp.2
Loranthus spp. (cf. L. micranthus)2
Compositae — tubuliflorae1
Weinmannia sp.1
?Astelia sp.1
unidentified monosulcate species5
Total count:230

Investigations have shown that other leaf floras from the South Island, ranging in age from Miocene to Eocene, exhibit a high percentage of leaf forms with entire margins but most were microphyllous in size. These floras may represent slightly cooler conditions than the one described above but more detailed investigations are needed, particularly of the associated pollen flora. An Upper Eocene sample from the Temuka Pottery Clay Pit contained a microphyllous flora with serrate/dentate leaves forming a large percentage.

I hope that these results will persuade taxonomists to give fuller descriptions of extant leaves enabling a more precise identification to be made of fossil leaves and thus allowing a more meaningful interpretation of palaeoclimate. It is possible that leaves of present day species were much larger in the warm phases of the upper part of the Tertiary and possibly Quaternary. Since this method has been used with great success overseas, particularly in America (Wolfe and Hopkins, 1967; Wolfe 1971), it is certain to be of value in New Zealand, where plants either had to migrate or change their habit during unfavourable periods in order to survive.

Experiment in the Identification of Plaster-cast Leaf Moulds

In order to discover if botanists can identify leaves without being able to see the growth form of the parent tree or shrub, or the colour, taste, feel and smell of their leaves, plaster-cast moulds of one hundred native New Zealand leaves were made and given to five botanists to identify. The botanists all had some experience in New Zealand taxonomy and ecology and were given unlimited page 82 time, and the use of books and herbaria for comparisons if needed.

The models were made by pressing a leaf into soft plaster-of-Paris, placing a strip of thin plastic on top and weighing the leaf down with additional plaster-of-Paris. Few leaves failed to form good impressions. After setting, the plaster-of-Paris easily separated along the ‘plane’ formed by the plastic.

The one hundred examples included some taxonomic replicas, i.e. different leaves from the same species. The results of this study are set out in the following table, arranged in taxonomic order after Allan (1961) to whom the reader is referred for author citations of the species.

o identified to genus

- identified to species

x incorrectly identified

A blank signifies that no attempt was made in identification
63Phymatodes diversifoliumX
54Asplenium lucidum-----
17A. flaccidum----X
42Podocarpus totara-----
50P. totara--o
32Dacrdium bidwilliiooXX-
6D. laxifolium-oXX
16D. cupressinum---
35Phyllocladus alpinus-o---
58P. trichomanoides-----
19Agathis australisXX
79Pseudowintera colorata
78Beilschmiedia tawa----
51Hedycarya arboreaX
43Laurelia novae-zelandiae-XX
61Clematis hookerianaoo-
11C. petrieiXo
88Ascarina lucidaXX
93Melicytus ramiflorus--X-X
10Hymenanthera obovataXX
18Fuchsia excorticata
65Pimelea longifoliaXX
15P. buxifoliaXX
100Knightia excelsa--XXX
26Coriaria arboreaoXX--
37Pittosporum tenuifoliumX-o-
24P. crassifoliumo--
99P. eugenioidespage 83
53Tetrapathaea tetrandraXX
70Metrosideros excelsaXo---
41M. robusta-o-
3M. fulgensX----
8Lophomyrtus bullata-----
64Eugenia maireX-
75Elaeocarpus dentatusXX
68Aristotelia serrata-XX
57Plagianthus betulinusXX
77Hoheria populneaX
44H. angustifoliaX
85Ackama rosaefoliaX
34Weinmannia silvicolaXoXX
36W. racemosaoX--
62Quintinia acutifoliaXX
52Ixerba brexioidesXX
21Carpodetus serratusXX
97Rubus cissoides--
30Clianthus puniceusXXX
86Sophora microphylla----
7Nothofagus menziesii-oo
38N. truncata---X-
39N. solandri-----
71Corynocarpus laevigatus
95C. laevigatusXo-X
9Pennantia corymbosao--
66P. corymbosaX--
59Elytranthe colensoiXX
89Pomaderris kumeraho-X
84Melicope ternataX-
74Dysoxylum spectabileXX
12Alectryon excelsusX
56A. excelsusX
49Dodonaea viscosaXo-
94Schefflera digitata---XX
20Pseudopanax colensoiXX-
80P. arboreum-o
47P. colensoiXX
92Corokia buddleioidesXXX
87Griselinia littoraliso-
31Gaultheria antipodao-ooo
1G. colensoi-oooo
96Gymnelaea cunninghamii-ooo-
25Geniostoma ligustrifolium---
40Parsonsia heterophyllao--o-
46P. heterophyllaoooo
55P. heterophyllapage 84
67P. heterophyllaoX-o-
69Alseuosmia pusillaX
33Coprosma foetidissimaXoXoo
4C. tenuifoliaXo-
83C. australiso-o
81C. robusta--o
45Olearia avicenniaefoliaXX
48O. albidaX
73O. albidaXXX
27O. arborescensX
90O. arborescens
98O. raniX-Xo-
14Senecio bidwilliiX-XX
76Brachyglottis repandao
60Ourisia macrophyllaX
22Hebe strictaXo
29H. strictaXXo
2H. tetragonaXo---
13Rhabdothamnus solandriXXXX-
5Vitex lucensXXX
28V. lucensXXXX
82V. lucensX-XX
91V. lucensXXX-
72Avicennia resiniferaXXX
23Rhipogonum scandensX-X--
Percentages correct to genus31%32%36%40%39%

The botanists were reluctant to fully commit themselves into identifying many of the moulds, probably due to the amount of time that it would have involved. The plaster-cast models show far more and better accentuated characters than leaves fossilised and impressed under natural conditions, so that it would be easier to identify plaster-of-Paris models than fossilised leaves of the same species. Thus it is surprising that the five botanists could only identify 60% to genus and 55% to species, while individually no one scored more than 40% correct to genus. The result was even more surprising as the specimens were, with few exceptions, well-known, common, native species. Poor casts may account for the failure of all botanists to identify numbers 44 (Hoheria angustifolia) and 74 (Dysoxylum spectabile).

We can look at the results from the point of view of operator error and the difference in their approaches to the problem, and from the view of leaf variability and difficulty of identification. For example, operator 2 attempted to identify 38 specimens and made only six mistakes while the other operators attempted 62-70 specimens and mis-identified 23-30 of them. Quite obviously operator 2 is the more accurate worker even though not proficent page 85 at identifying the majority of leaves (or any more than the other workers who introduced a higher proportion of guess-work). Quite a lot of information could be gathered from the attempts at identification. For example, there was little trouble in recognising Nothofagus species (Fagaceae) but no operator identified any of the Escalloniaceae (Ixerba, Quintinia and Carpodetus) even to genus.

Disregarding the operator variability, the problem caused by taxonomic replicas and the fact that 70% of the first fifty moulds were correctly identified to genus and only 50% of the last fifty, possibly suggesting a growing disinterest, there are some points which may be noted. These are as follows:


The following common and distinctive species were not identified at all — Agathis australis, Brachyglottis repanda, Carpodetus serratus, Corynocarpus laevigatus, Elaeocarpus dentatus, Fuchsia excorticata, Hoheria populnea, Olearia arborescens, Phymatodes diversifolium, Pseudopanax lessonii and Pseudowintera colorata. Other common, characteristic leaves were only identified once or twice.


Only a few of the moulds, including Lophomyrtus bullata, Nothofagus solandri, Phyllocladus trichomanoides and Podocarpus totara, were identified by all botanists.


Vitex lucens was identified as Coprosma repens six times and as Dysoxylum spectabile twice, and identified correctly only twice out of a possible of twenty times.


Rhabdothamnus solandri was identified as Nothofagus four out of a possible five times.


Weinmannia silvicola was identified as Laurelia novae-zelandiae three out of a possible of five times.


The three genera Carpodetus, Ixerba and Quintinia of the family Escalloniaceae were not identified at all.


Podocarpus, Dacrydium and Phyllocladus species were relatively easy to distinguish.


Metrosideros, a common leaf fossil, was readily identified in most cases to species and is readily distinguished from species of Pittosporum which were poorly identified.


Nothofagus species were identified readily.

The following general conclusions were made on the identification of fossil leaves.


To be sure of an accurate identification a palaeobotanist needs not only vein patterns, leaf margins, shapes and sizes (preservable characters), but also, in some cases, non-preservable characters as well as the use of extensive herbaria and personal experience.


Leaves of different genera may seem identical when fossilised.


Conspecific leaves of slightly varying shapes and sizes when found as fossils can readily be attributed to different species or genera.


A large herbarium of native and overseas plants must be available locally for comparisons of fossil leaves. For New Zealand the following overseas collections are essential for accurate comparisons: page 86 European and American genera that have ancestors common as fossils (e.g. Fagus, Ficus); an extensive collection of Malayan, Polynesian and Australian genera (especially such genera as Acacia, Casuarina, Protea, Banksia, Dryandra, Nothofagus ‘brassi’ group, etc.); and a collection of ‘primitive’ genera still surviving today (including Cycadaceae, various pteridophytes, Magnoliaceae, Winteraceae, etc.).


The inevitable possibility of mis-identification means that the palaeobotanist is giving names not only to true genera and species but also to varieties of leaf form within species. This could account for the large numbers of species within a genus in many overseas Cretaceous beds, particularly those described at the end of last, and the beginning of this, century. It would be realistic to admit that macropalaeobotany deals solely with form taxa throughout the geological column as is customary in palynology. Obviously the variation between congeneric specimens is far greater than conspecific leaves and this increases the magnitude of the identification problem many times.

It may be profitable to repeat this experiment using new plaster-cast models and more controlled conditions to test both operator experience and recognition of variations in leaf form. If any senior botanists would like to participate in this experiment I would be keen to hear from them. If five or more are willing to be tested I will prepare new models and send them to the interested parties.


Thanks are extended to the five botanists who gave generously of their time of participate in the experiment described above. The help of Dr C. A. Fleming, Dr M. T. Te Punga, Mr G. H. Scott (of the N.Z. Geological Survey), Dr J. W. Dawson and Professor H. W. Wellman (of Victoria University of Wellington) was much appreciated. Thanks are also extended to Mr B. C. Waterhouse who collected the macroflora described above. Special thanks are due to Mr Christopher Clowes who made many of the leaf measurements used in this study. This study forms part of a M.Sc project undertaken in the Geology Department of Victoria University of Wellington under Professor H. W. Wellman.


Allan, H. H., 1961: Flora of New Zealand V. 1. Government Printer, Wellington, New Zealand.

Dawson, J. W.; Sneddon, B. V., 1969: The New Zealand rain forest: a comparison with tropical rain forest. Pacific Science 23 (2): 131-47.

Madler, K.; Straus, A., 1971: Ein System der Blattformen mit spezieller Anwendung fur die Bestimmung neogener Blattreste (Miozan und Pliozan.). Botanische Jahrbucher 90 (4): 562-74.

Manze, U., 1968: Die Nervaturdichte der Blatter als Hilfsmittel der Palaoklimatologie. Sonderveroffentlichungen des Geologischen Institutes der Universitat Koln 14.

Mildenhall, D. C., 1968: The fossil flora of the Pakawau Group, North-West Nelson, New Zealand. M.Sc. Thesis, Victoria University of Wellington, 1968.

Millener, L. H., 1960: Our Plant World. New Zealand Junior Encyclopaedia, New Zealand Education Foundation, Wellington, 1960, 1: 310-36.

Mouton, J. A., 1966/67: Sur la systematique foliare en paleobotanique. Societe Botanique de France 113 (9): 492-502.

Richards, P. W., 1952: The tropical rain forest. Cambridge University Press, 450 pp.

Wolfe, J. A., 1971: Tertiary climate fluctuations and methods of analysis of Tertiary floras. Palaeogeography, Palaeoclimatology, Palaeoecology, 9 (1): 27-57.

Wolfe, J. A.; Hopkins, D. M., 1967: Climate changes recorded by Tertiary land floras in northwestern North America. In: K. Hatai (ed.), Tertiary Correlations and Climatic Changes in the Pacific — Symposium, Pacific Science Congress, 11th, Tokyo, August-September, 1966, 25:67-76.

1 * extinct species