!"#$%&'()%#*+%,-'.'/0*.-"/1/ 23.-'(4/56%71881#)%9:%;)1.-<%=-')#/%>:%?'@"8)#**<%AB#*%C:%D"!3E1#<%D#B1+%=:%F"88<%9"880%2: ;-"G-"(+ ;'3(E"6%F1';E1"*E"<%?'8:%HI<%J':%KK%4D"E:<%KLLI5<%GG:%IMNOILP ,3Q81/-"+%Q06%2)"(1E#*%R*/.1.3."%'$%F1'8'@1E#8%;E1"*E"/ ;.#Q8"%ST!6% 2EE"//"+6%UVWXKWUXXL%KK6NP

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission.

JSTOR is a not-for-profit organization founded in 1995 to build trusted digital archives for scholarship. We work with the scholarly community to preserve their work and the materials they rely upon, and to build a common research platform that promotes the discovery and use of these resources. For more information about JSTOR, please contact [email protected]

American Institute of Biological Sciences is collaborating with JSTOR to digitize, preserve and extend access to BioScience.

Leaf Form and Photosynthesis Do leaf structure and orientation interact to regulate internal light

and carbon dioxide?

William K. Smith, Thomas C. Vogelmann, Evan H. DeLucia, David T. Bell, and Kelly A. Shepherd

M orphological and anatomi- cal features of plant leaves are commonly associated

with metabolic type (e.g., Kranz anatomy of C4 species), amount of sun exposure (e.g., sun and shade leaves), or water stress (e.g., xero- morphism). However, although the primary function of the leaf is to absorb and process sunlight and car- bon dioxide for photosynthesis, few structural features of leaves have been related mechanistically to these tasks. For example, it has been known for over a century that the internal anatomy of leaves is characterized by different cell layers (e.g., the pali- sade and spongy mesophyll) and that stomatal pores can be located on one or both sides of a leaf. Yet, only re- cently has any functional relationship between leaf form and photosynthetic performance been suggested.

A variety of ecological studies have correlated numerous leaf structural parameters with photosynthetic per- formance (e.g., Abrams and Kubiske 1990, 1994, Hinckley et al. 1989,

William K. Smith (e-mail: [email protected] edu) and Thomas C. Vogelmann (e-mail: [email protected]) are professors in the Department of Botany, University of Wyo- ming, Laramie, WY 82071-3165. Evan H. DeLucia (e-mail: [email protected]) is an associate professor in the Department of Plant Biology, University of Illinois, Ur- bana, IL 61801. David T. Bell (e-mail: [email protected]) is an associate professor and Kelly A. Shepherd is a re- search assistant in the Department of Botany, The University of Western Austra- lia, Nedlands, WA 6907 Australia. ? 1997 American Institute of Biological Sciences.

Terrestrial plants responded to the

amount of sunlight and stress in a given habitat by evolving

leaf structural properties in concert

with leaf orientational capabilities

Koike 1988, Reich et al. 1991, Walter 1973), but mechanistic evidence pointing to a complex influence of leaf structure on photosynthesis has been obtained only recently (Tera- shima and Hikosaka 1995, Vogel- mann et al. 1996a). A comprehen- sive synthesis of the functional significance of leaf structure, as re- lated to photosynthesis, has yet to be proposed. In addition, no studies have associated leaf structural char- acteristics with differences in leaf orientation relative to the Sun, de- spite the recognition that both structure and orientation can have dominant influences on whole-leaf photosynthesis.

In this article, we present a syn- thesis of current findings in ecology, physiology, and biochemistry that points to a fundamental relationship between the evolution of leaf form (structure and orientation) and pho-

tosynthetic performance. This rela- tionship includes a strong coupling between leaf structure and orienta- tion that is not documented in the literature and that has not been at- tributed to photosynthetic function. We describe field observations of correlations among leaf structural symmetry, leaf orientation, and the resulting amount of incident sun- light on both leaf surfaces. We also summarize physiological and bio- physical evidence of the impact of this structural symmetry on the cap- ture and processing of sunlight and carbon dioxide for photosynthesis. We propose that the evolution of leaf structural symmetry is based on leaf orientation and the regulation of incident sunlight and is driven by a common functional theme-maxi- mizing photosynthesis per unit leaf biomass by regulating light and car- bon dioxide gradients inside the leaf. Although differences in chloroplast abundance, physiology, and behav- ior at different locations across the mesophyll are also important to this central theme (e.g., Evans 1996, Terashima 1992, Terashima and Hikosaka 1995), these topics are not emphasized.

For a typical plant leaf, sunlight is incident on the upward-facing (adaxial) side, whereas carbon diox- ide uptake occurs predominately at the lower (abaxial) side, where most, if not all, of the leaf stomatal pores are found (Figure 1; Meidner and Mansfield 1986). Thus, whereas chloroplasts just beneath the upper epidermis of this leaf should experi- ence the highest light regimes, the

December 1997 785

Figure 1. A cross-sec- Incident sunlight tion of a typical leaf showing the opposing U er gradients of internal epidermis ? O? light and carbon diox- uepidermisal ide when sunlight is in- ' CO?i cident on the upper leaf surface and stomata are o present predominantly i 5 ) % on the lower surface. Two pairs of hypotheti- pongy cal curves are drawn: Lower c. one pair (dashed lines) epidermis Stomata shows strong gradients Photosynthetic that generate a narrow zone of overlap (indi- cated by small bracket) between high light and carbon dioxide, and another pair (solid lines) shows smaller gradients that generate a broader zone of overlap (large bracket) between high light and carbon dioxide. A broader zone of overlap would generate greater photosynthesis per unit leaf biomass, which may be a fundamental driving force in the evolution of leaf form (i.e., structure and orientation).

carbon dioxide concentration is high- est on the opposite side of the leaf, next to the lower epidermis. Steep, opposing gradients in light and in carbon dioxide would not seem to be optimal for maximizing photosyn- thetic efficiency across the entire thickness of the leaf (Figure 1). It seems logical that leaf form would have evolved so as to maximize photo- synthesis per unit leaf biomass in the face of these opposing internal gradi- ents of light and carbon dioxide.

Does leaf structure regulate internal light? Considerable evidence indicates that the structural properties of leaves (apart from changes in chloroplasts) may influence photosynthetic per- formance. Most of this evidence comes from observations (Terashima and Hikosaka 1995) that the shape of the light-response curve of photo- synthesis (i.e., the amount of carbon fixed per amount of light) can be altered by changing the angle of inci- dence of direct-beam light, the direc- tional composition of the incident light (i.e., whether the beam is dif- fuse or direct), and the type of leaf structure (i.e., whether it is asym- metric or symmetric). Experimen- tally disrupting the parallel rays of direct-beam light by using a light diffuser caused substantial alter- ations in the light response of photo- synthesis (DeLucia et al. 1991, Terashima 1989). Similar alterations in photosynthesis have been observed

when structurally asymmetric leaves, which naturally intercept direct sun- light only on one surface, are illumi- nated on the opposite side instead (e.g., Evans et al. 1993, Kirschbaum 1987, Poulson and DeLucia 1993, Terashima 1989).

Increasing evidence implicates the leaf surface and all of the major cell types within a leaf (i.e., epidermis, palisade, and spongy mesophyll) as influencing the capture and internal processing of absorbed sunlight (Vogelmann et al. 1996a). Moreover, orientational and corresponding structural effects may have strong influences on photosynthetic prop- erties. Chloroplast acclimation to altered light regimes appears unable to compensate entirely for alterations in natural light regimes or normal leaf optical properties.

Upper epidermis. Leaf surface struc- tures, such as epicuticular waxes and epidermal hairs, have been reported to affect whole-leaf photosynthesis due to alterations in absorbed sun- light. For example, high solar reflec- tance from pubescent leaves of desert broad-leaf species results in optimal leaf temperatures, reduced transpi- ration, and enhanced photosynthe- sis (Ehleringer and Werk 1986, Johnson 1975, Smith 1978). Also, the hydrophobic nature of leaf pu- bescence found in numerous species may prevent a water film from form- ing during dew and rainfall, a poten- tially large barrier to photosynthetic carbon dioxide exchange (Brewer and

Smith 1994, 1997, Smith and McClean 1989). However, this same water repulsion may also create a monolayer of small water droplets over the entire leaf surface. Because of the lensing effects of these water droplets, a highly variable sunlight pattern develops over the leaf sur- face, ranging from full shade to over 20 times full sun at focal points be- neath individual droplets (Brewer et al. 1991). In most species tested, a layer of leaf trichomes holds the dew droplets above the leaf surface, well beyond their focal distances, greatly reducing the potential damage of this focused sunlight to the photosyn- thetic system.

Another common feature of the leaf epidermis is their lens-like cells, which were originally thought to be involved in orienting the leaf toward the sun (Haberlandt 1914). More recently, however, it has become clear that these lens-like epidermal cells both collect and focus incident light into the leaf interior, possibly to en- hance photosynthesis (Bone et al. 1985, Lee 1986, Poulson and DeLucia 1993, Poulson and Vogelmann 1990). These findings also show that the geometry of individual epidermal cells may vary according to sunlight exposure. Spherical epidermal cells may be more beneficial in shaded environ- ments, adding a much greater ab- sorbing area, not only for the pre- dominant levels of less intense diffuse light, but also for the direct sunlight (sunflecks) that penetrate the canopy at low angles of incidence (Smith et al. 1989). In addition, spherical epi- dermal cells would focus light to the shallow depths that are necessary for these typically thinner shade leaves. In sunnier habitats, more elliptical epidermal cells would generate deeper focal points for a more even distribution of internal light through- out thicker leaves (Vogelmann et al. 1996a). Moreoever, any bending of incident, direct-beam sunlight by epi- dermal cells is important for length- ening photon path lengths inside the leaf and, thus, increasing the prob- ability for absorption by chloroplasts (Vogelmann et al. 1996b).

Mesophyll. The optical properties of cell layers inside leaves (i.e., the pali- sade and spongy mesophyll) also appear to regulate the internal distri-

BioScience Vol. 47 No. 11 786

bution of sunlight for enhanced pho- tosynthesis (Vogelmann 1993, Vogel- mann et al. 1996a). For example, the more columnar palisade cells typical of thick sun leaves act as light con- duits that propagate light deeper into the mesophyll (Figure 1), thus dis- tributing light more evenly through- out the leaf (Terashima 1989, Vogel- mann and Martin 1993). In addition, the cell walls of the spherical spongy mesophyll cells and the large frac- tion of air space in the leaf interior generate large quantities of scattered light, increasing light absorption by chloroplasts within the mesophyll (DeLucia et al. 1996). Overall, inter- nal light scattering within leaves gen- erates photon fluence levels three to four times greater than sunlight inci- dent on the leaf surface, enhancing the absorption of weakly absorbed wavelengths in particular (Vogel- mann 1993).

Lower epidermis. Another funda- mental influence of epidermal struc- ture on photosynthesis may result from leaf bicoloration, in which the leaf side that faces away from the sun is lighter in color than the leaf surface facing toward the sun. Bicoloration is especially common in species that occupy more shaded habitats (Smith 1981). Bicoloration could enhance "light-trapping" in the spongy mesophyll by providing a reflective surface on the internal side of the lower epidermis (Lin and Ehleringer 1983, Smith 1981, Woolley 1971). In these studies, re- moval of the lower epidermis of a bicolored leaf resulted in large in- creases in light transmittance. The reflective properties of the spongy mesophyll and of the inside of the lower epidermis are also important for increased light retention and ab- sorption in bicolored leaves (DeLucia and Nelson 1993, DeLucia et al. 1996).

Light and carbon dioxide gradients in leaves It is reasonable to expect leaf orien- tation and structure to interact so that high light areas inside a leaf are matched with high carbon dioxide concentrations. Otherwise, full pho- tosynthetic potential will not be achieved (Figure 1). Although sub-

stantial gradients in light do appear to form across the leaf mesophyll (Vogelmann et al. 1996a), with cor- responding effects on whole-leaf pho- tosynthesis, carbon dioxide levels inside leaves have not been mea- sured directly, and much less is known about their characteristics (Parkhurst 1994). However, rela- tively large gradients of carbon di- oxide across the mesophyll thickness have been estimated (Parkhurst 1978) using indirect methods that measure carbon dioxide exchange in whole leaves that are exposed to carrier gases infused from different sides of the leaf (Parkhurst and Mott 1990). Estimates of up to a 16 Pa pressure difference in internal carbon dioxide between opposite leaf sides have been reported for leaves with large, ex- perimental differences in ambient carbon dioxide concentrations be- tween the two leaf surfaces and nearly equal numbers of stomata on both sides of the leaf (Parkhurst et al. 1988). Actual gradients of carbon dioxide inside natural leaves may be less, although the common occur- rence of stomata on only one side of the leaf would enhance steeper gradi- ents that would be in opposition to the light gradient (Figure 1). Parkhurst (1994) concluded that intercellular gaseous diffusion is a substantial limi- tation to photosynthetic carbon di- oxide assimilation in the large num- ber of species that have thick leaves and stomata on the lower leaf sur- face only. To date, measurements of both light and carbon dioxide gradi- ents within the same leaf are not available for any plant species.

Although carbon dioxide gradi- ents have not been measured directly inside leaves, experiments using pulse dosages of labeled carbon dioxide, with subsequent paradermal section- ing and autoradiography, have shown variation in the location of carboxylation activity inside leaves (Nishio et al. 1993). Initial studies indicated that the internal light gra- dients of sun and shade leaves of spinach did not correspond to the carbon fixation gradient (Nishio et al. 1993). However, a subsequent study reported that light absorption profiles predicted from chlorophyll concen- tration gradients did match carbon dioxide fixation profiles measured within spinach leaves (Evans 1996),

although this study did not measure internal light and carbon dioxide.

Logically, photosynthesis could be maximized if chloroplasts were situ- ated at locations within the meso- phyll at which both light levels and carbon dioxide availability were op- timized by the appropriate combina- tion of leaf orientation and struc- ture. The observation that mesophyll cell surface area, chlorophyll con- centration, and photosynthetic ac- tivity per unit leaf thickness are not uniform across the leaf thickness in- dicates that certain strata of the leaf may experience an optimum overlap of the opposing light (from above) and carbon dioxide (from below) gradients (Terashima and Hikosaka 1995). Evaluation of the relation- ship among leaf thickness, stomatal distribution, and whole-leaf photo- synthesis could provide ecophysiologi- cal evidence for the importance of the overlap of light and carbon dioxide gradients inside the leaf.

The interaction of leaf orientation and structure If leaf orientation and structure do interact to regulate sunlight absorp- tion and distribution inside the leaf, then the structural asymmetry iden- tified above (e.g., epidermal lens cells and palisade cells beneath the upper leaf surface of horizontal leaves) should correspond to the quantity and type of sunlight incident on each leaf surface. The focusing capabili- ties of epidermal lens cells require direct-beam sunlight (diffuse light is poorly focused by any lens), whereas palisade cells, if they function to propagate light deeper into the leaf, should occur beneath the leaf sur- face with greatest incident light. If carbon dioxide is to be supplied ad- equately to the increased mesophyll cell area in sun leaves, then the cor- responding increase in leaf thickness should be accompanied by a more equal distribution of stomata on both leaf sides. However, few ecological studies have related the occurrence of these structural differences in leaf symmetry, thickness, and stomatal distribution with differences in inci- dent light between the two leaf sur- faces under natural field conditions.

One might also expect to find changes in leaf structure that would

December 1997 787

diminish light absorption when a plant is experiencing other sources of stress-that is, when light is not limiting but temperature, water, or nutrients may be. Numerous studies have documented the detrimental impact of high light on photosyn- thetic performance, especially when a plant is under stress from other environmental factors (Baker and Bowyer 1994). For example, one rarely observes leaves of any species oriented perpendicular to full sun- light, unless leaf temperatures are low and transpirational water is abundant (Smith 1978). High inci- dent sunlight will result in leaf wilt (midday wilt) even for plants whose roots are in water-saturated soil (Young and Smith 1980).

One of the best-documented ob- servations of ecological patterns in leaf structure, already mentioned above, is the ability of most species to develop sun leaves under high sunlight exposure (e.g., Boardman 1977, Hansen 1917). In general, sun leaves are smaller in dimension (at least width, if not also length) but greater in thickness (e.g., De Soyza and Kinkaid 1991, Johnson 1978, Nobel 1991, Smith 1978). This re- duced leaf dimension in sun leaves generates a significant increase in convective heat dissipation, an im- portant factor for plant survival in drier, high-sun habitats, where over- heating and high transpiration rates are detrimental (Gates 1980).

The greater leaf thickness charac- teristic of sun leaves results in a sub- stantial increase in mesophyll cell surface area for carbon dioxide ab- sorption, providing a structural mechanism for the observed increases in photosynthesis per unit leaf area, even though photosynthesis per unit leaf biomass may remain unchanged (Nobel 1980). A greater mesophyll cell area also generates greater wa- ter-use efficiency because of the sub- stantially greater impact on carbon dioxide uptake than transpirational water loss. For species native to the most sun exposed, stressful habitats (e.g., desert shrubs, subalpine and boreal conifer trees), smaller, thicker leaves become almost cylindrical, with a more inclined leaf orienta- tion. Similarly, photosynthetic stems commonly replace true leaves in ev- ergreen shrubs of hot deserts, and

the frequent appearance of species with leaf and stem succulence (e.g., cacti and euphorbs) are further ex- amples of the occurrence of cylindri- cal geometry in highly stressful habi- tats. (We address the functional significance of a cylindrical leaf form in terms of light and carbon dioxide processing for photosynthesis in the next section.)

Most terrestrial plant species with thin, laminar leaves have many more stomata on the lower side of the leaf than on the upper side (i.e., they are hypostomatous), although a signifi- cant fraction (including most grasses) have almost equal numbers of sto- mata on both leaf surfaces (i.e., they are amphistomatous; Meidner and Mansfield 1986). Only a few species with thin, laminar leaves have sto- mata exclusively on the upper leaf side (e.g., lily pads; Brewer and Smith 1995). Increased leaf thickness has been associated with a more equal number of stomata on both leaf sur- faces for numerous species and taxa (Parkhurst 1978). Mott and Michael- son (1991) reported that increased incident light generated an increase in both leaf thickness and the num- ber of stomata on the upper leaf surface in Ambrosia cordifolia. Hav- ing stomata on both sides of a thicker sun leaf may increase the supply of carbon dioxide to the mesophyll cells (Mott et al. 1982, Parkhurst 1994, Parkhurst and Mott 1990). These studies provide evidence that the presence of stomata on both leaf surfaces greatly enhances carbon di- oxide supply to the greater meso- phyll cell area found in thicker sun leaves, both of which may be neces- sary to support the greater photo- synthetic rates per unit leaf surface area. Thus, both stomatal distribu- tion and mesophyll cell area contrib- ute to the higher rates of photosyn- thesis in sun leaves.

In a recent study, leaf structural and orientational data were collected for numerous evergreen species from five communities in Western Austra- lia to evaluate possible associations between leaf structure and orienta- tion (Smith et al. in press). These communities occur along opposing gradients in annual rainfall and daily incident sunlight due to an increase in understory species in the more mesic communities. At the time of

sampling, the five communities were composed of a high diversity of ever- green species only, whose leaves must endure seasonal drought (Beard 1990, Pate and McComb 1982). Such stress "tolerators" may be particu- larly indicative of adaptive relation- ships between leaf form and func- tion (Fahn and Cutler 1992, Levitt 1980).

For the five Australian communi- ties, strong positive correlations oc- curred between total daily sunlight and the proportion of species in a given community with thicker leaves, more cylindrical leaves, an inclined leaf orientation, palisade cell layers on both leaf sides, and stomata on both leaf sides (Smith et al. in press). Also, the presence of palisade cell layers on both leaf sides was corre- lated more strongly with a lower ratio (top-to-bottom) of incident sun- light than with the total amount of sunlight incident on the upper leaf surface only. By contrast, the num- ber of species with distinctly bicol- ored leaves (with the top side darker than the bottom side) was greater in the more mesic, shaded communi- ties. Because these understory spe- cies also had typical shade leaf struc- ture, leaf bicoloration was strongly correlated with the thin, laminar leaf structure and horizontal leaf display. Similarly, leaf bicoloration was nearly ubiquitous in understory plants of the subalpine zone of the Rocky Mountains (Smith 1981).

Corresponding changes in leaf orientation and structure in response to seasonal changes in stress is an- other example of the strong interac- tion between leaf structure and ori- entation. For example, the numerous drought-deciduous species in the deserts of the southwestern United States develop large, ephemeral leaves with horizontal orientation soon after rainfall (Beatley 1974). As the soil dries, these initial leaves are replaced by smaller, more in- clined leaves. With increasing soil dryness, numerous species shed these leaves and only green stems remain, generating a more inclined arrange- ment of curved photosynthetic sur- faces within the crown. Smith and Nobel (1977, 1978) also reported that high incident light had the great- est effect on leaf morphology (e.g., size, thickness, pubescence) and

BioScience Vol. 47 No. 11 788

Table 1. Influence of incident sunlight and stress level of the habitat on leaf orientational and structural characteristics and on photosynthetic potential in 234 species (86 families) of native plants (sampled predominantly from five Western Australia communities). Modified slightly from Smith et al. 1997.

Environmental conditions High sun,a High sun, Low sun,a Low sun,

Leaf form low stressb high stressb low stress high stress

Orientation Horizontal; tracks the sun Vertical or cylindrical; Horizontal Horizontal avoids the sun

Top-to-bottom ratio of >3.5c <2.0 <3.5 2.5-3.5 incident light Thickness (mm) >600 400-600 <400 <300 Thickness-to-width ratio <0.1 >0.1 <0.1 >0.1 Morphology Large laminar broad-leaf Small and cylindrical Large laminar broad-leaf Small linear or laminar

broad-leaf Hypostomatous and Amphistomatous Hypostomatous Hypostomatous amphistomatousd Structures to protect No bicoloration Bicoloration Weak bicoloration abaxial stomata; no bicoloration

Anatomy Upper palisade layers Upper and lower Single or no palisade layer No palisade layer palisade layers

Maximum photo- 1 2 3 4 synthetic potentiale

aDaily incident sunlight values computed over a 12-hour day were considered "high" if photosynthetically active radiation (PAR) was over 40 mol * m-2 * d-~ (as measured by a horizontal sensor) and "low" if PAR was less than 10 mol * m-2 * d-1. bStress was considered "high" if annual precipitation was less than 7 cm and "low" if it was greater than 10 cm. cAll values indicated for each category are rounded off to the nearest significant figure (e.g., to the nearest 100 for leaf thickness). dLeaves were classified as hypostomatous if more than 70% of the total leaf stomata were on the leaf underside; otherwise, they were classified as amphistomatous. eRelative ranking: 1 is greatest and 4 is least.

anatomy (mesophyll cell surface area and palisade development) in several drought-deciduous shrubs. How- ever, high light and temperature com- bined with low water stress gener- ated the thickest leaves. Thus, sunlight exposure and the level of water stress all interacted to signifi- cantly influence leaf structure. Korner et al. (1989) came to similar conclusions about the effects of tem- perature and light on leaf structure in high-elevation plants of the Cen- tral Alps.

Table 1 and Figure 2 present a synthesis, based on four generalized permutations of sunlight exposure and stress level in a habitat, that associates leaf orientational and structural characteristics with pho- tosynthetic potential. Plant species that have leaves with the greatest photosynthetic capacity occur in high-light, low-stress situations and have corresponding orientational and structural features that generate high photosynthetic rates-that is, hori- zontal, thicker leaves with multiple palisade layers on the leaf side facing the sun, and a more equal number of stomata on both leaf sides. As sun- light and stress increase, leaf orienta- tion becomes more inclined, with re-

duced sunlight interception, whereas leaf structure becomes more symmetri- cal (e.g., palisade cells occur on both top and bottom of mesophyll). With excessive sunlight exposure and stress, leaves become cylindrical, and the resulting radial diffusion elimi- nates the need for asymmetry in in- ternal anatomy. For species adapted to low-light regimes (i.e., that have horizontal, thin leaves with no pali- sade cells, and stomata only on the leaf underside), photosynthetic po- tential is low (Table 1 and Figure 2). These differences in leaf structure and photosynthetic potential can change within the same plant or among plants of a given habitat, ac- cording to seasonal changes in sun- light exposure or stress.

Evolutionary perspective The simplest …