Size shape and other structural characteristics

Water Freedom System

Survive Global Water Shortages

Get Instant Access

Leaf size has classically been dealt with in descriptive vegetation ecology by reference to leaf size classes. The Danish plant ecologist Raun-kiaer developed an arbitrary classification of leaf sizes based upon a geometric series 9" with a base leaf size of 25 mm2 (Raunkiaer 1916, 1934). He found that a series based on 9 produced a system that lent itself more meaningfully to the description of vegetation types across the globe than one based on 10. Each of the six size classes originally recognised was given a name derived from Ancient Greek for leaf (^dM,ov, phyllon) with a prefix indicating relative size. Webb (1959), while investigating the tropical vegetation of Queensland, Australia, found that many of the forests he was studying fell into the mesophyll-dominated class but there was still a distinction in types. To provide a means of doing so, he divided the mesophyll class into two, recognising the notophyll class (22.5-45cm2) as the designation for the small mesophylls, with large mesophylls becoming the mesophyll class sensu Webb.

The mesophyll sensu Raunkiaer leaf size is the commonest among tree species in lowland tropical rain forests (Table 2.5). Usually at least three quarters of species belong to this group. Another feature of tropical lowland forests is the presence of at least a few species that reach the megaphyll class (> 1640.25 cm2), even if palms are omitted from the analysis.

Rain-forest tree leaves are typically entire-margined (Bailey & Sinnott 1916; Rollet 1990) with an ovate-lanceolate shape, roughly three times as long as broad (Bongers & Popma 1990a) and frequently with a pronounced acuminate tip. Compound leaves are usually quite common; for example, of 183 species surveyed in Venezuela, 57 (31%) had compound leaves (Rollet 1990).

Table 2.5. Leaf sizes in various lowland tropical rain forests

Values given are the percentage of the species sampled in each of the leaf size classes of Raunkiaer.

Table 2.5. Leaf sizes in various lowland tropical rain forests

Values given are the percentage of the species sampled in each of the leaf size classes of Raunkiaer.

Site

lepto

nano

micro

meso

macro

mega

Mexico1

0

3

9

79

8

1

Ecuador2

0

0

9

64

27

0

Brazil3

2

2

13

74

7

0

Brazil3

3

1

12

75

9

0

Brazil4

1

0

15

78

5

1

Nigeria5

0

0

10

84

6

0

Gabon6

2

6

11

80

1

0

Gabon6

1

3

8

84

4

0

Gabon6"

0

0

9

82

5

0

Gabon6

1

4

8

86

1

0

Philippines7

0

0

4

86

10

0

Sources: Mongers et al. (1988); 2Grubb et al. (1963); 3Cain et al. (1956); 4Mori et al. (1983); 5Richards (1996); 6Reitsma (1988); 7Brown (1919). "Data, as given in source, do not sum to 100%.

Sources: Mongers et al. (1988); 2Grubb et al. (1963); 3Cain et al. (1956); 4Mori et al. (1983); 5Richards (1996); 6Reitsma (1988); 7Brown (1919). "Data, as given in source, do not sum to 100%.

The relative rarity of toothed leaves and other non-entire margin forms in the tropical rain forest is well enough established for palaeobotanists to use the frequency of toothing in leaf fossil assemblages, together with leaf size, to infer the type of climate under which the vegetation represented by the assemblage grew (Wiemann et al. 1998). There are two basic angiosperm leaf venation patterns: craspedodromous, where secondary veins form a pinnate arrangement and run more or less parallel to each other and terminate at, or near, the leaf margin, and brochidodromous, where the secondaries loop and join within the leaf margin (Fig. 2.17). Leaves of temperate deciduous trees are frequently craspedodromous with a toothed margin, but those of tropical rain-forest trees are more likely to be brochidodromous with an entire margin. Roth et al. (1995) employed a hydrodynamic model to develop hypotheses concerning the relative merits and constraints of the different leaf designs. The advantage of the craspedodromous pattern over the brochidodromous one is that it uses a shorter length of vein per unit leaf area and therefore would make leaves cheaper to construct if 'plumbing costs' are high and should make leaves more efficient at intercepting light as a greater proportion of the lamina area can be devoted to photosynthetic tissues. The hy-drodynamic model also highlighted the limitations of the craspedodromous design. The delivery of water to cells farthest away from the terminal ends of the veins, i.e. areas on the leaf margin between the secondaries, was poor. This probably explains the toothed leaf margins of many craspedodromous leaves. The tissues on the leaf margin that would be vulnerable to dehydration or over-heating because of the low rate of water supply are simply not formed,

Figure 2.17 Schematic representation of the two basic leaf venation patterns of dicotyledons.

resulting in a toothed edge. Rain-forest leaves generally live longer than do those of temperate deciduous forest (Coley & Barone 1996), so therefore it is probably necessary to invest in a more reliable water-supply system for the mesophyll tissues, represented by brochidodromous venation.

The acumen of tropical leaves is often referred to as a drip-tip because of the tendency of water to accumulate there, forming an ever larger drop until it drips off. There has been considerable speculation about the advantages of having a means of increasing the rate of drainage of water from the surface of a leaf (Richards 1996). These include reduced periods of a water layer obscuring the incoming radiation and leaching minerals, particularly potassium, from the leaf tissues, and the likelihood of water encouraging the growth of epiphylls. A shrubby species of Piper in the montane forests of Costa Rica was shown to drain drops of water from leaves at a slower rate when the drip-tip was removed (Lightbody 1985). However, no allowance was made for drop volume. Drip-tips appear most pronounced on the leaves of tree saplings growing in the shade (Roth 1996) and in species with pinnate leaves (Rollet 1990).

Along gradients of increasing seasonality of rainfall there is usually an increased proportion of drought-deciduous species and leaf size tends to decrease (Gentry 1969). Increasing altitude is generally reflected in a decrease in relative abundance of large-leaved species (Dolph & Dilcher 1980). Tree leaf size, frequency of compound leaves and frequency of drip-tips was found to decrease with elevation in New Guinea (Grubb & Stevens 1985).

There is considerable variation both within and among tropical forests and forest types in leaf structural characteristics such as leaf mass per unit area (LMA), lamina thickness, anatomical features and concentrations of nutrients (Figs. 2.18 and 2.19). The collation of data from published reports on leaf properties does not show such a clear trend as is generally described in the literature. The expectation is for the leaves of species from higher or more infertile sites to be smaller and thicker, with more dry mass per unit area, and probably volume, and to have lower concentrations of important nutrients (Grubb 1977; Turner 1994).

The typical leaves of lowland heath and upper montane forests in the tropics can be referred to as sclerophylls. The name means 'hard leaf', and is an allusion to the textural properties of the leaf in comparison to the soft and flexible leaves of more mesic sites. Leaf texture includes properties such as hardness, stiffness, strength and toughness. The first two, resistance to impression and bending, do not have any successful methods for quantification in leaves. Tensile strength and work of fracture (toughness) can be measured by using calibrated load cells in the appropriate equipment (Lucas & Pereira 1990; Lucas et al. 1991a). The highly anisotropic nature of leaf material causes difficulties. Frequently the midrib and veins are tougher than the intercostal region of the lamina (Lucas et al. 1991a; Choong et al. 1992). It is not easy to derive an average toughness or strength for the leaf, and in practice pre-determined fracture paths have to be chosen for comparative purposes.

The average LMA of different forest types varies considerably among sites (Fig. 2.18), with a large degree of overlap among the different forest formations. A more direct study of leaf form on one mountain, Gunung Silam in Borneo, has shown a trend of increasing LMA with altitude (Bruijnzeel et al. 1993). LMA also increased down a soil fertility gradient at San Carlos de Rio Negro in Venezuela (Medina et al. 1990). Grubb (1998a) reported a positive correlation between LMA and leaf size for species from caatinga in Venezuela

Vensuleas Traits

Figure 2.18 Species average leaf characteristics for different forest sites. Circle, mesophytic lowland forest; triangles, lowland heath forest; squares, montane forest. NPPR, non-palisade to palisade tissue thickness ratio in the mesophyll. Data from Bongers & Popma (1990a), Choong et al. (1992), Turner et al. (2000), Sobrado & Medina (1980), Tanner & Kapos (l982), Kapelle & Leal (1996), Sugden (1985), Cavelier & Goldstein (1989), Coomes & Grubb (1996), Grubb (1974) and Peace & MacDonald (1981).

Figure 2.18 Species average leaf characteristics for different forest sites. Circle, mesophytic lowland forest; triangles, lowland heath forest; squares, montane forest. NPPR, non-palisade to palisade tissue thickness ratio in the mesophyll. Data from Bongers & Popma (1990a), Choong et al. (1992), Turner et al. (2000), Sobrado & Medina (1980), Tanner & Kapos (l982), Kapelle & Leal (1996), Sugden (1985), Cavelier & Goldstein (1989), Coomes & Grubb (1996), Grubb (1974) and Peace & MacDonald (1981).

and argued that this reflected a requirement for proportionally more material for support in larger leaves. However, this correlation is not evident in data provided by Rollet (1990) or Turner & Tan (1991). At Los Tuxtlas, Mexico, there was a positive correlation between leaf area and mass-based concentrations of N, P and K (Bongers & Popma 1990a) across the species in the lowland forest.

Lamina thickness more convincingly distinguishes the supposedly sclero-morphic types from more mesophytic lowland forest (Fig. 2.18). However, non-palisade to palisade tissue thickness ratios in the mesophyll (NPPR) do not appear to show any consistency in distinguishing forest types. Montane forests tend to have a lower density of stomata than lowland ones (Fig. 2.18). The canopy leaves of mesophytic lowland forests can be as tough as those of heath forests (Turner et al. 1993b, 2000). Heath and montane forest species do appear to have generally lower foliar concentrations of major nutrients (Fig. 2.19). This is most noticeable for nitrogen where the ranges of community averages for mesophytic lowland and heath forests barely overlap, but for most of the other elements the heath and montane forest averages rarely approach 60% of the maximum value of a mesic lowland site. There was relatively little difference in mass-based foliar nutrient concentrations among a small sample of arborescent dicot and monocot species at La Selva (Bigelow 1993). Area-based concentrations were higher in the monocots; this reflected the higher LMA of these species.

Was this article helpful?

0 0

Post a comment