Plant Identification

by Anna Lawrence

  130 Plant Identification

  •

  If your users expect your guide to work for saplings, do not underestimate how much fieldwork it takes to record the sapling details for compound leaves accurately and to modify any herbarium-based description. You will have to identify the saplings by finding and linking all stages between sapling and adult. Try not to base keys on leaflet numbers when saplings have to be identified as well. Other features such as glands, venation and hairs are usually more constant.

  •

  In bipinnate leaves with many small leaflets (for example, Acacia, Piptadeniastrum and Parkia), the individual leaflets or even pinnae often vary little as the plant grows up (the leaflets are often only slightly larger in the shade); but the total number of pinnae and size of leaf typically varies more.

  •

  Check carefully whether the leaflets on a pinnate leaf are precisely paired and opposite or sometimes slightly unpaired. The closely paired arrangement with no terminal leaflet (absolutely paripinnate) occurs more commonly in legumes than the type of ‘paripinnate’ leaf where the leaflets are not quite opposite (that is, sub-opposite).

  •

  Be careful to indicate what type of ‘imparipinnate’ leaf you mean when specifying pinnate leaves with some unpaired leaflets – are the leaflets precisely paired with one unpaired leaflet on the end (paired + 1), almost paired +1, or alternate? Species with sub-opposite leaflets common on herbarium specimens are frequently found in nature to possess juvenile or shade leaves where the leaflets are unpaired to the point of being alternate (or with only one leaflet).

  •

  The terminology associated with bipinnate or tripinnate leaves where lobing occurs on some leaflets and not others is complicated; but these characteristics are easy to demonstrate with silhouette-type diagrams, easily made with a scanner and easily converted to small monochrome diagrams (see Figure 8.1).

  •

  In other cases, it is useful to illustrate large compound leaves with a full view of the whole leaf at one scale and a single leaflet with more detail shown closer to the camera or artist.

  Leaf shape and margin

  Characters such as leaf shape and margin are amply catered for by virtually all botany textbooks, dictionaries, Floras and so on (see Figure 6.1). Rather few words are routinely used to describe the shape of a leaf, and although it is generally much more desirable to include a picture of leaf shape than words, a few shape terms are useful in even a non-technical field guide – for example, in a guide to an incomplete set of species

  – to describe a rarer unillustrated relative that differs in leaf shape. However, even these terms are used inconsistently (see Box 6.6). Software developers have not yet achieved an ideal way of indexing shape, although this is likely to happen soon – and this will revolutionize pictorial e-keys (see Box 6.7).

  Marginal details such as teeth and lobes are very important in most field guides. For a useful manual primarily for workers on fossil leaves see the Leaf Architecture Working Group (LAWG, 1999), where it is pointed out that the most reliable characters of a leaf are entire versus serrated margins, lobing, and primary and secondary venation. For non-technical field guides, it is enough to distinguish lobed, serrated (toothed) and crenate from entire margins; but one can specify much finer detail, preferably with illustrations, to distinguish subtly different types of tooth for species that are hard to distinguish.

  Plant characters suitable for field guides 131

  BOX 6.6 WHAT EXACTLY IS EGG-SHAPED?

  Consider the words ovate (egg-shaped) and lanceolate (lance or spear-head shaped). Both of these have been defined for botanists (Systematics Association, 1962; followed by Stearn, 1966, and many others) as shapes broadest below the middle, the ovate with a length/breadth ratio of between 2:1 and 3:2 – in other words, shapes between twice as long as wide and one and half times as long as wide. Alternatives such as (very) widely ovate can approach 1:1 – that is, almost round but still broadest before the middle. Lanceolate leaves are much more slender, at 6:1 to 3:1. This still leaves open to question what to call a shape between 6:3 (= 2:1) and 6:2 (= 3:1). More serious is the fact that many people still use the ambiguous ‘egg-shaped’ definition, while accepting some very dubious shapes of eggs.

  Compare the Flora of Australia glossary distinction between ovate and lanceolate (see www.anbg.gov.au/glossary/webpubl/splitgls.htm): this is significantly different from Harris and Harris (2000), whose illustration at least looks egg-shaped. Furthermore, some botanists seem to ignore obovate and oblanceolate, where the broadest point of the leaf is near the apex, choosing instead to use the words ovate or lanceolate no matter which way round the leaf is. We strongly recommend using the ‘ob-’ forms for leaf shapes in field guides, and to follow the standards for the numeric limits (see Figure 6.1).

  There are many aspects of leaves that reflect a plant’s ecology, guild or functional group: for instance, leaf size, inclination from horizontal, serrated margins, cordate bases, drip tips and many more traits are loosely associated with ecology (for a summary, see Givnish, 1987, and Press, 1999). By choosing to emphasize some of these characters in a key, it is often possible to distinguish meaningful ecological groups – for example, where single pages have similar leaves from similar habitats to ease comparison (Gillison, 2002)

  Leaf stalks

  Leaf stalk is a non-scientific term for petiole, but more familiar to the public and therefore often appropriate in simpler guides:

  •

  Beware of the very common confusion between the terms petiole, petiolule and rachis (see Box 6.4).

  Petioles provide many useful field characters in themselves – for example, if swollen at one or both ends (‘pulvinate’), channelled, winged, many-grooved, twisted, articulated, hairy or twining. But they are also useful at providing an easily locatable, standard part of a leaf for characters that tend to vary from one part of the leaf to another, such as hairs and glands:

  •

  The hairs that occur on petioles are often little different from the range found elsewhere; but it is often worth specifying the indumentum specifically on petioles (defined positions on the midrib or along the apical bud are also useful) where there is variation across a plant.

  132 Plant Identification

  Moderate levels of jargon in a field guide glossary Figure 6.1

  Hawthorne (1990); Hawthorne et al (2005)

  Source:

  Plant characters suitable for field guides 133

  BOX 6.7 DEFINING AND ANALYSING LEAF SHAPES IN DATABASES

  How might shapes be represented in a database so that they can be matched, perhaps automatically, to a specimen to be identified? There are no off-the-shelf solutions today, so should field guide writers give this a moment’s thought? If you want to be at the cutting edge of the subject, maybe you can organize your data to make the most of such developments when they arrive. The problem of subtle shape-detection software is a mathematical one with biological constraints (see, for example, Ledig et al, 1969; Dale et al, 1971), and various approaches have been proposed and compared (see Freeman, 1961; Dickinson et al, 1987; Jensen, 1990; Loncaric, 1998). Some programmes and algorithms are still under development, or recently reported but far from commercially available or reliable (Abbasi et al, 1997; Liu and Sclarof, 2000; Gandhi, 2003). It is not difficult to match exact shapes, but generally inconvenient to codify a leaf to be identified, especially in the field, and to deal with acceptable variation and cases where, for instance, a spectrum of possibilities exists for one species from, say, three to five lobed to unlobed leaves:

  •

  A useful set of recommendations for basic leaf measurements to include in your database is given by the Leaf Architecture Working Group (LAWG, 1999), starting with midrib length multiplied by lamina width (for lobed leaves, draw an ellipse joining the lobe tips and treat this as a leaf), and distance along midrib of greatest width. Such a simple shape index can be measured and typed i
n manually (LAWG, 1999) and used to shortlist species in a database, but has much lower discriminatory power than is theoretically possible (for example, based on matching by eye).

  •

  Database searches for words such as ‘ovate’ or for the length/breath ratios that define them may be of some use for narrowing options, but are far from subtle unless the shape is very unusual. However, the rarest shapes have no precise word to describe them, the ratios are likely to miss their essential oddness, and these species are the easiest to identify from pictures anyway.

  •

  Do not spend long typing in complicated shape data; soon software will, one assumes, be available to scan images in databases and to extract indexable details.

  Rather, start now to build up your digital picture library of clean, isolated leaves on a plain, unpatterned (black or white) background. Such images are also ideal for orthodox use in a field guide. By contrast, most images of herbarium specimens or living plants show too much overlap of leaves, ripped or ragged margins, perspective distortion or three dimensional ‘artefacts’, such as shadows on the mounting paper, and are unlikely to be easily searchable or indexable for a number of years (see also O’Callaghan, 1970, and Ray, 1992).

  •

  Many petioles have clear ‘abscission lines’, or lines of weakness marking where the leaf will break off when their time is up – this is a more reliable and useful character than referring to a plant as ‘deciduous’, although it is not necessarily the case that the correlation is absolute.

  •

  The maximum petiole length is a useful character to record for all species; emphasize to your users that it is important to look across many stems of a plant to find this.

  •

  When describing the size of leaves, be sure to explain where the leaf length is measured from and to. Many authors do not, and in these cases the length is presumably of the petiole and lamina (leaf blade) combined. It is usually more useful to divide total leaf length into ranges for the length of the petiole and length and width of the fully developed leaf blade or lamina.

  134 Plant Identification

  •

  Sometimes it is not actually the absolute length of the petiole that is significant – for instance, for reducing leaf overlap – but the length relative to the lamina.

  Petiole/lamina length ratios can sometimes, therefore, be useful in a key.

  Some of the characteristics of petioles apply to petiolules (leaflet stalks in compound leaves); for instance, abscission joints, with an abrupt change of texture or colour between petiolule and rachis, are common at the base of petiolules of legumes and various vegetatively similar families. In many legumes, the petiolule is often short, fat and wrinkled like an earthworm, or else a so-called ‘cushion’ where a leaflet blade with an asymmetric base is hinged directly onto the rachis. These can be useful features for distinguishing them from other pinnate-leaved groups. The petiolules of the leaflets of compound leaves are usually short (< 1cm long) and not pulvinate; but in Dacryodes and other Burseraceae they may become longer or swollen at the top.

  Finer details of the leaf blade (lamina)

  Leaf texture

  There are various ways of describing the obvious variation in thickness of the lamina of fresh leaves, especially leathery (= coriaceous), papery (= papyraceous) and even thinner than paper, such as many of the more delicate fern leaves (= membranaceous). This is very far from an exact science, and although the (bracketed) terms are important in descriptions, they can often be ignored in favour of other characters in keys. For instance, in leathery leaves, margins are often recurved, and midrib or vein prominence may be quite different from close relatives with different leaf blade thicknesses. Between leathery and papery, some leaves have the texture of a thin sheet of brittle plastic, crack-ing when folded. The term ‘plastic textured’ can be used for these, the ancient Romans not having invented plastic.

  There is scope to develop a mechanical leaf texture index to use with a field guide as the character is potentially more useful than the jargon allows for. Perhaps you could bind flaps of different types of material in your book or actual dried leaves for a herbarium guide?

  Leaf colour

  Some leaves, particularly of ornamentals, are variegated or strikingly coloured, making them very distinctive in pictorial guides. For non-cultivated plants, a distinction between otherwise similar species could be made in the field on the basis of shades of green, associated with whether the plant is evergreen or the leaf is leathery. However, it is difficult to record this accurately, even on camera, and to print it out successfully; often, one has to be satisfied with mentioning extremes only. However, for difficult groups, leaf surface colour and related details visible only with a good lens might be useful in technical guides (see, for example, Stace, 1965).

  For a tree or liane guide, the best source of leaves is the forest floor, and the fallen leaves of many species turn a distinctive colour – for instance, fallen Calophyllum calaba leaves are usually mottled red and green, and are usually to be found below an adult. In the African swamp species Spondianthus preussii, lacy leaf skeletons are always found below the tree. The distinctive colours of dried specimens may also be useful for identifying them in the herbarium (for example, the reddish, black or greyish shades in various

  Plant characters suitable for field guides 135

  Note: Leaf venation, often so distinctive, is best illustrated. An easy method is to photograph fresh leaves against the light, converting if necessary to black and white.

  Source: William Hawthorne

  Figure 6.2 Leaf venation

  Psychotria species); but be careful because different methods of drying can result in different colours.

  Glaucous leaf surfaces, often pale blue or white due to a wavy layer, are very distinctive, often diagnostic and survive drying well.

  Leaf venation

  The basic patterns of venation are best expressed by illustration, especially photographic (see Figure 6.2), although there are some terms that are easily understood without illustrations and others that are useful for emphasizing features that may otherwise not be noticed (Hickey, 1973; Hickey and Wolfe, 1975, although their jargon is not very user friendly). For many descriptions, it is important to refer to the venation order (1°, primary, for the largest order of veins; 2° for laterals; and so on). Although the casual definition of vein order seems simple enough, problems arise when trying to be absolutely precise – for example, when comparing closely related species with subtly different venation, or when specifying venation patterns accurately in databases (see Hill, 1980; Spicer, 1986):

  •

  The minimum and maximum number of pairs of lateral nerves (2° venation) is a very useful character to record for all species.

  •

  Illustrate venation by photographing a leaf with a macro lens, held up to a bright light. This can be manipulated to form a monochrome diagram-like image (see Figure 6.2 and Chapter 8).

  •

  It is usually possible to pick on some arbitrary aspect of the venation when only two species are compared – for instance, if the finest venation looks more lax in one species than another, try comparing the number of areoles (finest, closed vein fields) crossed on a line between two adjacent laterals in the middle of the leaf, halfway to

  136 Plant Identification

  the margin. These sorts of indexes can be surprisingly constant. The angle of lateral nerves to the midrib is sometimes useful; but few field botanists carry a protractor, so try to use this character for difficult groups or extreme differences only.

  Midrib and vein prominence

  There is much variation in the conspicuousness of veins, partly due to their width, partly due to their prominence (above or below the lamina surface), and partly due to how opaque the surface of the leaf is, possibly hiding finer veins. 2° and finer veins may be

  ‘obscure’, meaning not clearly visible. It is important to specify whether th
is applies to the upper or lower leaf surface, and what minimum order of venation it applies to (for example, 3° venation partly obscure; 4° and finer venation completely so). Is the appearance something that is only apparent at arm’s length, or are all finer veins invisible on close inspection, as well?

  The degree of prominence of the midrib on the upper surface of the leaf is a significant character in some groups and largely survives drying well, so the information can be gleaned from herbarium specimens very quickly. It is possible to go beyond the three main character states of prominent, channelled and flat (see Figure 6.1) and to distinguish ‘impressed’ from ‘channelled’ and ‘guttered’ from ‘prominent’. For the midrib, it is usually best to specify a part of the leaf where this character is to be checked; the midrib, for instance, may be channelled where it runs into the petiole and prominent halfway along the leaf.

  For finer venation (3° and higher order), the degree of prominence on either surface can vary with drying; in particular, finer venation that is prominent on the upper surface of dried leaves is usually less so in the field (although fallen leaves may be dried in this respect). The degree of prominence on the lower surface is more constant, and veins that are impressed, flat or guttered on either surface in fresh leaves tend to remain so when dried.

  Smells, taste and bark: The gourmet field botanist Smell and taste, whether of tree barks or fresh leaves, are often useful clues to a plant’s identity (see Box 6.8). Obviously, fleshy fruits are frequently sweet or sharp and pleasant tasting and smelling, and nuts and leaves are sometimes edible and tasty. More intriguing and less seasonal, though, is the range of other testable or scented plant parts. Tasty leaves are more common among savanna herbs and shrubs than they are among rainforest woody plants; but a few forest species have distinctive tastes. Some Dialium species (oxalic acid taste, like rhubarb) and Tamarind leaves have a pleasant, sharply acidic and fruity taste, useful for distinguishing them from otherwise similar legume saplings. The leaves of some Begonia species are tasty (sour) and palatable enough to be collected as a salad vegetable.