Crop Science Journal of Natural Resources and Life Sciences Education
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
 QUICK SEARCH:   [advanced]


     


Published in Crop Sci 39:1584-1596 (1999)
© 1999 Crop Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (36)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Loomis, R.S.
Right arrow Articles by Amthor, J.S.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Loomis, R.S.
Right arrow Articles by Amthor, J.S.
Agricola
Right arrow Articles by Loomis, R.S.
Right arrow Articles by Amthor, J.S.
Crop Science 39:1584-1596 (1999)
© 1999 Crop Science Society of America

SYMPOSIUM-1998 ASA MEETING -BALTIMORE

Yield Potential, Plant Assimilatory Capacity, and Metabolic Efficiencies

R.S. Loomisa and J.S. Amthorb

a Agronomy & Range Science, Univ. of California, Davis, CA 95616, USA
b Environmental Sciences Division, Oak Ridge National Lab., Oak Ridge, TN 37831-6422 USA

rsloomis{at}ucdavis.edu


    ABSTRACT
 TOP
 ABSTRACT
 INTRODUCTION
 Photosynthetic Production
 Respiration and Biosynthesis
 Assessments
 REFERENCES
 
The structure, control, and efficiency of photosynthetic and respiratory systems are examined. Genetic control is complex and highly conserved. While many features are still unresolved, basic efficiency seems little altered by domestication and breeding of crops. Rubisco, the carboxylase–oxygenase enzyme central to photosynthesis and photorespiration, remains a weak point but may be amenable to improvement. However, the actual radiation-use efficiency of crops is generally less than the potential with present rubisco kinetics, leaving considerable room for improvement without change in rubisco. Good opportunities for progress lie in definition of optimal canopies of leaves having suitable acclimation and photoprotection. The efficiency of the respiratory system also seems unaffected by plant breeding. Precise evaluation of the roles and efficiencies of the glycolytic pathway and the tricarboxylic acid cycle in production is difficult because, in addition to being sources of energy carriers and reductant, those systems also supply carbon skeletons for biosyntheses. How those systems are controlled and balanced for such diversions is largely unknown. The alternative oxidase found in mitochondria may be involved in that balance but its true role(s) is also unknown. Distinguishing two components of respiration, one related to maintenance and the other to growth, remains a powerful tool in theoretical studies. In such work, the respiratory system appears efficient, but proving that in experiments remains elusive.

Abbreviations: AMP, ADP, and ATP, adenosine mono-, di-, and triphosphate, respectively • C3 plants, plants that employ only the reductive pentose phosphate cycle in photosynthesis • C4 plants, plants that employ PEP carboxylase to concentrate CO2 as the first step in photosynthesis • CH2O represents a unit of carbohydrate with MW = 30 g mol-1 • LAI, leaf area index • NADP+ and NADP(H), oxidized and reduced nicotinamide adenine dinucleotide phosphate, respectively • PAR, photosynthetically active radiation • PEP, phosphoenolpyruvate • PFD, photon flux density • PSI and PSII, photosystems I and II, respectively • RuP2, ribulose-1,5-bisphosphate • RUE, radiation-use efficiency • SB, SM, and SR, substrate carbon (or carbohydrate) retained in new biomass, or expended in maintenance or biosynthesis, respectively • TCA, tricarboxylic acid


    INTRODUCTION
 TOP
 ABSTRACT
 INTRODUCTION
 Photosynthetic Production
 Respiration and Biosynthesis
 Assessments
 REFERENCES
 
PLANT PRODUCTION

is driven by photosynthesis. Key elements in the system are (i) the interception of photosynthetically active radiation (PAR, 400–700 nm spectral band), (ii) use of that energy in the reduction of CO2 and other substrates (photosynthesis), (iii) incorporation of the assimilates into new plant structures (biosynthesis and growth), and (iv) maintenance of the plant as a living unit. Achieving high yield is conceptually simple—maximize the extent and duration of radiation interception; use the captured energy in efficient photosynthesis; partition the new assimilates in ways that provide optimal proportions of leaf, stem, root, and reproductive structures; and maintain those at minimum cost.

In detail, the processes are complex. The dynamic pattern of partitioning is particularly important. Crop yield comprises only a portion of biomass that accumulates over a crop cycle. Effective root and canopy systems (including stem structure for foliage display), for example, generally must be established before the onset of reproductive effort. In addition, the cost of maintenance increases as vegetative biomass accumulates during the season. Because crops are at the mercy of spatial and temporal variations in weather, plant spacing, and supplies of water and nutrients, and in occurrence of pests and disease, flexibility in morphogenesis and acclimation of the physiological systems is a key requirement for achieving high and stable performance.

With limited space, we focus on photosynthesis and on respiration related to synthesis and maintenance. Whether the biological efficiency of these processes has, or might be, improved through breeding are important questions.


    Photosynthetic Production
 TOP
 ABSTRACT
 INTRODUCTION
 Photosynthetic Production
 Respiration and Biosynthesis
 Assessments
 REFERENCES
 
Whether a canopy (amount of leaf area, LAI, and its manner of display) is optimal for photosynthesis in a particular environment is reciprocally linked with development and properties of individual leaves, including their longevity. According to a leaf's position in the canopy, variations occur in the components of its photosynthetic system, its acclimation to changing conditions, and its protection from excess photon flux density (PFD).

Leaf Components
The solar-energy-capturing apparatus of higher plants is located in thylakoid membranes of chloroplasts. As summarized in Fig. 1 , it consists of light-harvesting antennae complexes composed of carotenoids and chlorophylls a and b connected to Photosystem (PS) I and II reaction centers, a cytochrome b6f complex, and ATP synthase. The b6f complex transfers electrons from PSII, the water-oxidizing center, to PSI leading to NADP+ reduction. The proton gradient that develops across the thylakoid between an interior lumen and the exterior stroma is employed by ATP synthase (coupling factor complex, CFo-CF1) to produce ATP from ADP.



View larger version (18K):
[in this window]
[in a new window]
 
Fig. 1 Light reactions of photosynthesis and photophosphorylation associated with thylakoid membranes inside chloroplasts leading to production of NADPH and ATP used in CO2 assimilation. Photons are absorbed by chlorophyll antennae in both photosystem II (PSII) and photosystem I (PSI). Water is oxidized by PSII, yielding O2 and protons in the lumen. Coordinated activity of PSII and the cytochrome (cyt) b6f complex pumps protons from stroma to lumen. This involves cycling of plastoquinone between oxidized (Q) and reduced (QH2) states. As drawn, a "Q-cycle" is associated with cyt b6f. Electrons are transferred from cyt b6f to PSI via plastocyanin (PC). In noncyclic electron transport, e- are then transferred from PSI via ferredoxin (Fd) to FNR (ferredoxin-NADP reductase), which leads to reduction of NADP. The dashed line (– – –) indicates cyclic transport in which e- flow from PSI back to the cyt complex via Fd. Protons move from lumen to stroma through CFo-CF1 ATP synthase, which catalyzes ADP phosphorylation (about 3 protons/ATP)

 
The photosynthetic reductive pentose phosphate cycle ("dark reactions" involving CO2 assimilation) is found in the stromal solution. The key enzyme, rubisco, catalyzes both the oxygenation and carboxylation of ribulose-1,5-bisphosphate (RuP2). Rubisco's activity as an oxygenase, the initial step in the process of photorespiration, increases as the ratio [O2]/[CO2] at the enzyme and/or temperature increase. Also in the stroma are enzyme systems that manufacture and repair chloroplast constituents, reduce nitrite and sulfite, and synthesize starch. The scene is further complicated by the presence of the glycolytic system in the stroma (as well as in cytosol) and by exchange of various compounds between cytosol and chloroplast. For example, triose phosphate moves to cytosol, where sucrose synthesis occurs, in exchange for inorganic phosphate, Pi. Any slowing of this Pi recycling, or accumulation of end products, can slow photosynthesis. The Q10 of the reductive cycle is near 2 and low temperature limits CO2 reduction unless the capacity is increased through increases in enzyme concentrations.

All chloroplasts of C3 plants contain the full set of enzymes for CO2 assimilation. In C4 crop plants, rubisco and most of the carbon reduction cycle occur only in chloroplasts of bundle sheath cells. C4 mesophyll cells, by contrast, lack rubisco but rely on an important cytosolic enzyme, phosphoenolpyruvate (PEP) carboxylase, to assimilate CO2. In C4 plants of the NADP-malic enzyme type (maize, Zea mays L., and sugarcane, Saccharinum spp.), PEP carboxylase fixes CO2 into oxaloacetate, a four-carbon organic acid, which is then reduced to malate. Malate is transferred to bundle sheath cells where it is decarboxylated to pyruvate, thus concentrating the dilute supply of CO2 around rubisco and greatly reducing photorespiration. The pyruvate returns to mesophyll chloroplasts where it is converted to PEP through conversion of ATP to AMP. The cycle is completed when the PEP returns to mesophyll cytosol. An additional complexity in C4 plants is that 3-phosphoglycerate (PGA) is also exported from bundle sheath cells and is reduced in mesophyll chloroplasts, thus utilizing reducing power that is available there.

Genetic Control of Photosynthesis
Considerable progress has been made towards understanding the molecular biology of photosynthetic systems (Ort and Yocum, 1996). Thylakoid and carbon reduction systems comprise well over 100 proteins, and the stroma and outer membranes of chloroplasts are sites for an array of proteins related to transport, coordination, and regulation of chloroplast activity. Hankamer et al. (1997) list 27 genes and protein subunits solely for PSII. Genes for most chloroplast proteins are known and sequencing is well advanced. That work offers evidence that the PSI and PSII reaction centers were derived some 2 x 109 years ago from several endosymbioses with photosynthetic bacteria. The principal source was cyanobacteria and as many as five endosymbioses may have occurred (Reith, 1996). Genetic control over synthesis of these proteins is divided between nuclear and chloroplastic DNA. Chloroplast DNA, which is circular with a high level of ploidy (20–900 copies per chloroplast), may code for 100 proteins (Erickson, 1996). Additional problems for genetic manipulation arise from the maternal cytoplasmic (nonnuclear) inheritance of chloroplasts (and mitochondria) and the large numbers of these organelles in higher plant cells.

C4 species evolved at various times (10–15 x 106 years ago) from many different C3 species resulting in numerous variations in the system. The CO2-concentrating action and anatomical modifications greatly improve adaptation to high temperatures and to low atmospheric [CO2], conditions that accelerate photorespiration in C3 plants. Other benefits include better water-use efficiency than C3 species and, perhaps more important, need for much less rubisco (and thus less N) per unit leaf area for rapid photosynthesis. Genetic control is similar to that for C3 systems, and key enzymes are homologous with ancient C3 genes, but the genes are expressed differently (Monson, 1999).

There is no evidence that important beneficial changes in the structure of thylakoid and CO2-reduction-cycle components occurred during domestication of crop plants or more recently through breeding. Properties of most components are highly "conserved" in the sense that very little genetic variation is found among higher plants and, indeed, between higher plants and their ancestral green algae (Chlorophyta).

An obvious target for genetic engineering is to improve rubisco's affinity for CO2 thus reducing photorespiration (Long, 1998) and, perhaps, need for a high concentration of this enzyme. With an improved enzyme, the amount of N per square meter of leaf area needed to achieve a given photosynthesis rate might be less because rubisco accounts for a large fraction of the N in C3 leaves. Considerable genetic variation is found in rubisco—not surprising considering its very large molecular weight (>500 000)—and the enzyme's properties vary within and among crop species (e.g., Makino et al., 1988). Rubiscos from C4 species have higher Km (Michaelis-Menton constant) for CO2 (i.e., less affinity for CO2) than those from C3 species (Yeoh et al., 1980, 1981). The high Km values are correlated with greater catalytic capacity of the C4 carboxylases and less inhibition by high [CO2] (Yeoh et al., 1980). Yeoh et al. (1980) speculated that a tradeoff exists between capacity and affinity. It may be that reduction in C3 photorespiration through greater rubisco affinity for CO2 would be linked with a lower catalytic capacity, and thus a need for greater amounts of the enzyme and N in leaves.

Much remains to be done in identifying roles for many of the proteins found in thylakoids. Given the complexity of the thylakoid system and its genetic control, advances through genetic engineering may come slowly, if at all. Progress seems more likely through changes in rubisco, in leaf acclimation within the canopy, and in protection from excessive excitation of chlorophyll. Many of the advances in yield achieved during the past 50 yr have come through improvements in canopies. In maize, for example, large increases in plant density coupled with much greater N supply and "stay green" character resulted in more rapid achievement and greater duration of full cover. Further progress will require better understanding of what constitutes an optimal canopy. Acclimation and protection are central issues in that question and deserve special attention.

Acclimation
Crop plants are exposed to widely fluctuating conditions of light and temperature, and supplies of water and nutrients, and have evolved with a leaf-level photosynthetic apparatus that is highly flexible in structure and activity. Depending upon environment, leaves develop with different numbers and sizes of cells, different numbers of chloroplasts per cell, and with variations in amounts and proportions of thylakoid and carbon-reduction-cycle components. Changes in these factors seem more related to photosynthetic activity per unit C and N invested in leaf structure than per unit leaf area. Acclimative changes depend on light environment and position in the canopy and continue on a time scale of days to weeks throughout the life of a leaf.

Acclimation to light is proportional to mean daily irradiance of the leaf rather than to peak irradiance (Chabot et al., 1979). This ability is important because new leaves generally emerge at the top of a crop in full sun and later are submerged into the shade of the canopy as other leaves develop above them. C3 leaves in full sun typically have more of their leaf N involved in electron transport and carbon reduction and less in light harvesting (and fewer grana stacks) than is the case for shade leaves (Terashima and Evans, 1988; Evans, 1993; Pons and Pearcy, 1994). These properties also vary with depth within the leaf from its sunlit surface (Evans, 1995; Terashima and Hikosaka, 1995). In Evans' (1993) study of alfalfa (Medicago sativa L.) canopies, soluble proteins (e.g., rubisco) declined more with depth in the canopy than did thylakoid proteins. In addition, chlorophyll a/b decreased with increasing depth in the canopy, reflecting a decline in reaction centers relative to light-harvesting antennae. These changes resulted in a decline in chloroplast N/chlorophyll in a way that maintained photosynthetic capacity per unit N, a key trait for optimal distribution of N within the canopy.

Leaf adjustments to limiting supplies of N are especially important, involving changes in number and size of new leaves as well as in proportions of thylakoid and carbon-reduction-cycle components with depth in the canopy. Optimal distribution of N among leaves within a canopy is important and has received attention in recent years (Dreccer et al., 1998). The problem also involves canopy architecture, solar track, sky condition, and time remaining in the season (Loomis, 1993). For young crops with small leaf area, increasing leaf area for greater radiation interception provides more benefit than increasing photosynthetic capacity of existing leaves (through greater N content per unit leaf area).

Protection
Leaves exposed to full sun encounter challenges in balancing electron transport with their capacity for carbon reduction and/or the supply of CO2. For crop plants well supplied with water and nutrients, present low atmospheric CO2 concentration (near 360 µmol mol-1 air) is a greater problem than reduction capacity. Daily photosynthesis of such crops can be equivalent to all of the CO2 in 100 m of air. Although atmospheric turbulence extends the mixing to much greater heights and ensures that concentrations within the canopy generally remain above 250 µmol CO2 mol-1 air, CO2 concentration within leaves (Ci) limits maximum photosynthesis rates in C3 crops. At the heart of the situation is rubisco's low affinity for CO2.

As light flux increases and Ci becomes limiting to the carbon reduction cycle (Fig. 2) , the light-response curve of CO2 uptake departs from a linear increase (minimum number of quanta required per CO2 reduced) and plateaus at a maximum, light-saturated rate (Amax). Absorption of light energy continues but the excitation energy cannot be dissipated in the usual way because ADP and NADP+ substrates are not available to accept electrons; i.e., the reaction centers are "closed". Limitations to CO2 supply because of stomatal closure (e.g., drought stress) or low capacity for CO2 reduction (e.g., N deficiency or low temperature) increase the likelihood that PFD will be in excess of rate of use.



View larger version (25K):
[in this window]
[in a new window]
 
Fig. 2 CO2 uptake (PS rate) by a leaf with increasing photon flux density (PFD). Departure from the initial slope (which defines the minimum quantum requirement) occurs as CO2 becomes limiting and photosynthesis approaches a maximum, light-saturated rate (Amax). The shaded area represents what Björkman and Demmig-Adams (1994) term "excess PFD"

 
Closure places reaction centers at risk of short- or long-term photodamage. The D1 protein in the core of PSII is particularly vulnerable given the large redox potential (about 1.17 eV) developed there and the presence of singlet oxygen (1O2) (Barber, 1998). As a result, this protein turns over very rapidly (once or twice per hour) in illuminated leaves. Repair is expensive: the protein must be ejected from the PSII center, rebuilt, and reinserted. Such photodamage increases the minimum quantum requirement and reduces Amax.

With small leaf area, canopy photosynthesis is maximum with spreading leaves despite high levels of light saturation and the potential for photodamage. With more leaf area, canopy architectures that result in less irradiance per unit leaf area (e.g., erect leaves) are effective in limiting these problems. Some crops, notably legumes, alter leaf display during the day in ways that reduce light absorption in high light (reviewed by Koller, 1990) and, in all crops, small portions of the excess energy may be expended in nitrite or sulfite reduction or in leaf maintenance.

In C3 leaves, the reaction centers may be kept open and excess energy dissipated by utilization of NADPH and ATP in photorespiration. In this, rubisco catalyzes the condensation of O2 with RuP2, rather than CO2, and glycolate (a potentially toxic compound) and PGA are generated. In terrestrial plants, glycolate is metabolized in peroxisomes and mitochondria (where CO2 is released), producing glycerate. Glycerate is taken up by chloroplasts where it and PGA participate in regeneration of the RuP2 substrate used by rubisco. In this way, three of every four C atoms are recycled allowing another cycle of O2 or CO2 reduction unless serine (which is involved in a mitochondrion-to-peroxisome transfer) is "drained" off to protein synthesis.

Unfortunately, photorespiration increases with temperature because of decreased solubility of CO2 in the stroma and decreased affinity of rubisco for CO2 relative to O2 (i.e., decreased specificity for CO2). Even in dim light, when there is no need for protection, photorespiration increases the quantum requirement for CO2 reduction (qr) of C3 plants from a minimum value near 11 mol photons mol-1 CO2 at low temperature (<15°C) to about 25 at temperatures near 35°C (Ehleringer and Björkman, 1977). The slow rate of photorespiration at low temperature means that it cannot offer much protection against over-excitation under those conditions. Björkman and Demmig-Adams (1994) offered data on performance of cotton (Gossypium hirsutum L.) (Table 1) as evidence that photorespiration also provides only modest protection under drought.


View this table:
[in this window]
[in a new window]
 
Table 1 Carbon dioxide assimilation (PS-CO2) and photorespiration (PR-O2) per unit leaf area and their contributions to dissipation of chlorophyll excitation energy in cotton leaves under full sun with variations in water supply. Leaves of well-watered plants ({Psi}leaf = -1.0 MPa) were compared with leaves from plants under extended drought. Adapted from Björkman and Demmig-Adams (1994)

 
To what extent rubisco acts as carboxylase or oxygenase depends upon the relative concentrations of CO2 and O2 presented to the enzyme. Under elevated atmospheric CO2, the CO2 concentration within C3 leaves increases and oxygenase activity is suppressed. C4 plants suppress photorespiration in ambient air by the same principle—the CO2-concentrating action of their mesophyll cells keeps rubisco (in bundle sheath cells) well supplied with CO2. Oxygenase activity is then only a small percentage (2–6%) of the net CO2 flux (de Veau and Burris, 1989).

Osmond and Grace (1995) considered another form of protection based in Mehler and ascorbate-peroxidase reactions, which operate well in water-stressed plants, as perhaps more important than photorespiration for protection of C3 leaves. This involves formation at PSII of peroxide (H2O2), which is then reduced by ascorbate peroxidase. The oxidized ascorbate is regenerated by a reductase with expenditure of NADPH. These reactions keep both reaction centers open without a net exchange of O2. Because study of this pathway depends on examining oxygen isotope discrimination, which is difficult, little is yet known about its protective role.

Another system of protection exists in interconversion of carotenoid pigments found in association with chlorophyll in light-harvesting complexes (Demmig-Adams and Adams, 1996; Gilmore, 1997). With high PFD, pH changes in the thylakoid lumen induce conversion of violaxanthin to zeaxanthin (Fig. 3) , which can accept resonance energy from chlorophyll. When reaction centers close, excitation energy passes from chlorophyll to zeaxanthin and is converted to thermal energy (increasing leaf temperature) before it reaches the PSII reaction center. The leaf thermal energy is dissipated to the environment through convection, transpiration, and long-wave (infrared) radiation emission. Such "nonphotochemical quenching" of excited chlorophyll is assayed easily through measurements of variable chlorophyll fluorescence. In dim light, zeaxanthin recycles to violaxanthin, which cannot intercept energy from chlorophyll. Björkman and Demmig-Adams (1994) and Gilmore (1997) view nonphotochemical quenching as a major mechanism for protection of PSII from excess PFD.



View larger version (20K):
[in this window]
[in a new window]
 
Fig. 3 Daily course of the xanthophyll photoprotective system in cotton under field conditions (adapted from Björkman and Demmig-Adams, 1994). (Top) Incident PFD, efficiency of PSII, and fraction of PSII centers that were closed. This amount of closure is much less than would occur without intervention of the xanthophyll cycle. (Bottom) Concomitant changes in concentrations of xanthophyll pigments during the day as violaxanthin was converted to zeaxanthin, which is capable of accepting excitation energy from chlorophyll

 
Little is yet known about the need for photoprotection by crops. Young crops with spreading leaves are at the most risk. Within canopies, leaves are displayed at angles to the sun's rays that greatly limit the amounts of excess PFD they absorb. In addition, most crop plants are "sun" plants with a large capacity for photosynthesis and less need for photoprotection than plants having less capacity. Sun plants generally have high stomatal conductance that helps maintain Ci and allows sunlit leaves to dissipate heat by transpiring freely. Yield advance in CIMMYT wheat (Triticum aestivum L. var. aestivum) lines, for example, has been associated with increased stomatal conductance (Fischer et al., 1998). That trait obviously is most useful for well-watered conditions.

Debate continues on the relative roles and merits of various protective systems. Osmond and Grace (1995) viewed photorespiration and the ascorbate-peroxidase reactions as the main protection. Andrews and Baker (1997) suggested that photorespiration buffers against imbalances in PFD absorption by PSI and PSII. Björkman and Demmig-Adams (1994) and Long (1998), by contrast, saw little merit in photorespiration and view leaf acclimation, canopy architecture and the xanthophyll cycle as the principal means of protection. Long (1998) suggested that photorespiration persists only because evolution has reached a "barrier" for improvements in rubisco's affinity for CO2. He noted that rubisco in some Rhodophyta has greater affinity for CO2 and might serve as a source of genetic material.

Radiation-Use Efficiency
Crop growth rate and yield are functions of canopy photosynthesis and they generally correlate poorly if at all with maximum photosynthesis rates of individual leaves. Given the oblique display and mutual shading of leaves within canopies, few leaves are exposed to PFD sufficient to achieve Amax. Other reasons for the discrepancy can be found in acclimation to radiation level, temperature and stress, and the amount of standing crop (and thus the amount of maintenance respiration). As a result, crop physiologists have sought other measures that would relate yield and canopy photosynthesis. Light-conversion efficiency (commonly termed radiation-use efficiency, RUE) has received the most attention. RUE is measured and reported in various units, e.g., g new biomass produced MJ-1 radiation intercepted or absorbed by leaves. A useful feature of RUE is that experimental values can be compared with estimates of potential rates of dry matter production that might be possible by a canopy of well-acclimated leaves. Potential RUE would be attained with all leaves exposed to only moderate PFD (little or no light saturation and, thus, minimum qr). If that crop at the same time intercepted most of the incoming radiation, its rate of biomass production per unit land area would also be maximized.

We recently reexamined potential RUE for C3 plants in light of modern understanding of quantum requirements and C losses in respiration (Loomis and Amthor, 1996). Calculations were done for a C3 crop with 1000 g m-2 standing biomass (the midseason amount for a crop producing 10 Mg grain ha-1), moderate maintenance respiration, and a growth yield in biosynthesis (YG) of 0.72 g new biomass g-1 assimilate consumed. (YG is discussed below in the section on a Functional Model of Respiration.) Calculated RUE varied from 4.1 g MJ-1 solar radiation absorbed for qr = 10 to only 1.1 at qr = 30. This wide range of qr embraced the large increase in photorespiration that occurs in C3 leaves with increasing temperature (Ehleringer and Björkman, 1977). By these calculations, the average performance observed for high-yielding wheat crops (near 1.5 g MJ-1 solar radiation absorbed, with roots included; Fischer, 1983) corresponded to a qr of 24 (Loomis and Amthor, 1996). Some wheat crops reached 2 g MJ-1, however, corresponding to qr = 14.

Similar calculations for maize-type C4 species are presented in Table 2 . Given the small amount of photorespiration in C4 leaves, only a small range of qr values around 16 is considered. The minimum possible qr for C4 plants is larger than for C3 species because ATP is expended in malate production, a cost that is increased by leakage of concentrated CO2 from the bundle sheath leading to "over-cycling" of malate (Jenkins, 1997). Ehleringer and Pearcy (1983) reported minimum values for several C4 species in a narrow range around qr = 15.4 mol photons mol-1 CO2. That high efficiency seems to require participation of the thylakoid Q-cycle (enhanced ATP production) to offset the cost of malate over-cycling (Jenkins, 1997). With qr = 16, 1400 g biomass m-2, and YG = 0.74, the calculation presented in Table 2 predicts an RUE of 2.3 g MJ-1 solar radiation intercepted (4.6 g MJ-1 PAR intercepted).


View this table:
[in this window]
[in a new window]
 
Table 2 Estimates of potential radiation-use efficiency (RUE) of C4 crops at different quantum requirements. Assumptions are for a maize crop with a closed canopy completely intercepting incident solar radiation and with all leaves functioning at a uniformly small quantum requirement. For convenience, calculations are based on 1 MJ of incoming solar radiation (with PAR = 0.5 total solar) and all values scale linearly with the level of incoming radiation. CH2O is a unit of carbohydrate with MW = 30

 
Calculated RUE values are sensitive not only to variation in qr (which varies with radiation level; see Fig. 2) but also to variation in maintenance respiration, which depends on temperature, maintenance coefficient, and size of standing crop. For the system presented in Table 2 with qr = 16, variations in maintenance lead to the following:

Maintenance (mmol CH2O):0 20 40 60 RUE(g MJ-1 solar): 2.7 2.4 2.0 1.5

where CH2O represents carbohydrate with MW = 30 and maintenance is per megajoule solar radiation intercepted. RUE is also sensitive to variation in YG, which declines as concentrations of protein and/or lipid increase; i.e., the amount of assimilate used per gram new biomass increases. Thus, the largest RUE occurs for crops with a large content of carbohydrates (cellulosic material, starch, sugars). Where comparisons are to be made between crops differing in composition, RUE should be expressed either in glucose equivalents required in synthesis (see below; and Flenet and Kiniry, 1995) or in energy content of the biomass. At qr = 16, RUE of 2.3 g biomass MJ-1 solar radiation intercepted can be expressed as equivalent to 104/6 = 17.3 mmol glucose MJ-1 (Table 2). Biomass of this composition would have a heat of combustion near 17.6 kJ g-1 corresponding to over 4% of absorbed solar radiation (and near 9% of absorbed PAR).

Gallo et al. (1993) and Sinclair and Muchow (1999) discussed difficulties inherent in experimental determination of RUE (large SEs for both growth rate and radiation interception) and offered protocols for good results. There have been reports of seemingly valid measurements of 4 to 5 g MJ-1 PAR intercepted by healthy maize crops during vegetative growth but Sinclair and Muchow rejected those larger than 3.2 to 3.4 g MJ-1 PAR as failing to meet their standards. Their maximum values can be increased by about 10% to 3.5 to 3.7 g MJ-1 PAR to account for roots but they still fall well short of the 4.6 g MJ-1 PAR estimated in Table 2. As was found with calculated RUE values, measured values for C4 crops generally exceed those of C3 crops, and, for maize, smaller values were found during cool weather (Andrade et al., 1993) and under stress, such as high evaporative demand (Kiniry, 1999). In one of the few successful comparisons of cultivars, Tollenaar and Aguilera (1992) found that postanthesis RUE of a maize hybrid released in 1988 exceeded that of a 1959 hybrid. The difference was related to the "stay-green" trait possessed by the 1988 hybrid.

Contrary to the hopes of crop physiologists, the variable nature of RUE and difficulties in its measurement prevent it from serving, as a sensitive measure for exploring the fine structure of photosynthetic systems. Uncertainty about what constitutes a "reliable" value of RUE also raises serious questions about its wide use as a driving variable for crop simulation models. Perhaps the RUE era should be closed. It persists, however, because the alternatives for study of canopy photosynthesis are difficult or expensive and because crop modelers resist replacing it by submodels that simulate canopy photosynthesis.


    Respiration and Biosynthesis
 TOP
 ABSTRACT
 INTRODUCTION
 Photosynthetic Production
 Respiration and Biosynthesis
 Assessments
 REFERENCES
 
The Respiratory System
Respiration in higher plants is commonly viewed as a sequence of enzymic steps. With hexose as a generic substrate, flow of carbon can be traced through the glycolytic pathway (found in cytosol and plastids) to the tricarboxylic acid (TCA) cycle in the matrix solution of mitochondria. Mitochondria, like chloroplasts, are enclosed by an outer membrane that encompasses a convoluted inner membrane, inside which is the matrix (Fig. 4) . The basic scheme of respiration is that hexose (e.g., glucose) is partially oxidized to pyruvate, which is then taken up by mitochondria where it is completely oxidized to CO2. A portion of the energy released is captured in reduced nucleotides (NADH and FADH2), which may in turn transfer their protons (H+) and associated e- to a series of protein complexes located in and on the inner mitochondrial membrane. These complexes serve in electron and proton transport. Protons are moved to the intermembrane space, creating a gradient across the inner membrane that serves to drive Fo-F1 ATP synthases (Boyer, 1997) that convert ADP and Pi to ATP. The H+ and e- are eventually reunited in reduction of O2 to H2O.



View larger version (25K):
[in this window]
[in a new window]
 
Fig. 4 Respiratory-chain reactions associated with the inner mitochondrial membrane. Mitochondria are enclosed by two membranes; the outer membrane (not shown) is permeable to metabolites but the inner one is not. NAD(P)H in the mitochondrial matrix and in the cytosol can be oxidized by several mitochondrial dehydrogenases. Complex I oxidizes matrix NADH and in so doing pumps protons from the matrix to the cytosolic side of the inner membrane. Other inner-membrane-bound dehydrogenases (indicated by DH) do not pump protons. Inner-membrane dehydrogenases (including complex II, which functions as part of the TCA cycle) transfer electrons (and protons) to ubiquinone (Q), which is reduced to ubiquinol (QH2) in the process. A mobile Q/QH2 pool exists in the inner membrane. Complex III oxidizes QH2 and passes electrons to cytochrome (cyt) c, which in turn passes electrons to complex IV. Electron transport through complexes III and IV is coupled to proton translocation across the inner membrane; a "Q-cycle" associated with Complex III is included in the diagram. Free O2 is reduced to water by complex IV. The alternative oxidase (alt ox) can also oxidize QH2, and form water, but this bypasses two sites (complexes III and IV) of proton translocation in the mitochondrial electron transport chain. Protons may enter the matrix through membrane leaks, but Fo-F1 ATP synthase (similar to the CFo-CF1 ATP synthase in Fig. 1) is the main route of proton entry. Apparently, one ADP is phosphorylated when three protons pass through the F0-F1 ATP synthase. The ADP required for ATP formation enters the mitochondrial matrix only as ATP exits the matrix through an antiporter. A symporter couples the transport of Pi and H+ into the matrix

 
Several variations may occur in this pathway. The oxidative pentose pathway, for example, routes some of the glucose-6-P from the beginning of the glycolytic pathway through pentose sugars to fructose-6-P and glyceraldehyde-3-P (glycolytic intermediates). CO2 is liberated and NADP+ is reduced. Possible functions of this oxidative pentose phosphate pathway, in addition to production of NADPH, are production of ribose-5-P and erythrose-4-P, which are important precursors in biosyntheses. Another alternative route occurs at the end of glycolysis where a "malate shunt" (Amthor, 1994a) may bypass pyruvate kinase through production of malate from PEP by PEP carboxylase. The shunt is favored when ADP is in short supply. Malate serves as substrate in amino synthesis and can be stored in vacuoles (balanced by K+, it serves there as an important osmoticum) or metabolized in the TCA cycle. An interesting point is that pyrophosphate (PPi), rather than ATP, can be used in phosphorylating hexose at the start of the glycolytic scheme and this may be an essential process in some plants (Plaxton, 1996).

The products of glycolysis (pyruvate and malate) can be completely oxidized in the TCA cycle with production of ATP and reduced nucleotides. As detailed in Fig. 4, the great bulk of ATP production then occurs through oxidation of the nucleotides by protein complexes located in the mitochondrial inner membranes. If all reducing agents produced by glycolysis and the TCA cycle are employed in ATP production, a total of about 30 mol ATP mol-1 glucose can be produced (Amthor, 1994a; Stryer, 1995). (This amount is less than 36 mol ATP commonly quoted in older biochemistry texts.) About half of the free energy of hexose is captured in ATP when a hexose molecule is completely oxidized in respiration; the rest is lost as heat. Most of the "retained" energy is also lost as heat when ATP is subsequently used (hydrolyzed). Thus, when hexose is completely used for the formation of ATP, nearly all its free energy is eventually lost as heat.

Alternative Oxidase
The alternative oxidase found in mitochondria (alt ox in Fig. 4) deserves special comment. Electrons that pass to this oxidase bypass two sites (complexes III and IV) of proton translocation with the result that less ATP is formed. Lambers et al. (1998)(p. 106–111) summarize possible roles for the oxidase. Most relevant to competitively grown crops is the theory that it serves to waste protons and e- when ATP is already in abundance, allowing the TCA cycle and glycolysis to proceed in production of carbon intermediates needed in biosyntheses. The oxidase is under a high degree of coarse and fine control (including induction by organic acids and rapid activation), however, indicating that its role may be more fundamental (Vanlerberghe and McIntosh, 1997).

The magnitude of e- flux through the alternative oxidase remains uncertain but we found no evidence that it has a significant impact on crop performance. Early studies were done with cyanide (inhibits e- transport to complex IV) and SHAM (salicylhydroxamic acid; inhibits the alternative oxidase) but these have side effects that affect respiratory rates. Better information is obtained with noninvasive isotope methods (alternative oxidase discriminates against heavy O2 more than complex IV; Guy et al., 1989). Little alternative oxidase activity was observed in rapidly growing roots with the isotope method, whereas more than half the total e- flux in older roots terminated at the alternative oxidase (Millar et al., 1998).

System Nature and Control of Respiration
It is important to emphasize the system nature of respiration. Many of the steps are reversible and may be employed, for example, in synthesis of sugars from lower-level compounds. In addition, rather free exchange of various intermediates (e.g., hexose phosphates, dihydoxyacetone-P, and PEP) occurs between the cytosolic and plastid glycolytic schemes. Particularly significant is the diversion of glycolytic intermediates to synthetic pathways: PEP to shikimic acid pathways to various alkaloids and lignin; dihydroxyacetone and acetyl CoA to lipid synthesis; and acetyl CoA to isoprenoids (carotenoids, abscisic acid).

The TCA cycle is also a factory for intermediates with oxoglutarate serving as precursor to porphyrins (chlorophyll, cytochrome) and amino acids (and thus protein). As a result, respiration rarely (perhaps only in some nongrowing tissues) operates in the simple linear series outlined earlier. This remarkable flexibility is important in balancing widely varying demands for ATP, reductant, and intermediates found in various tissues and at various times. The ability to employ reducing agents directly in syntheses or in production of ATP is particularly important. The oxidative pentose phosphate pathway and the malate shunt also contribute flexibility (and complexity).

The flexibility of the system depends upon coarse control at several levels with the PEP-to-pyruvate conversion (pyruvate kinase) being one key. Except in rapidly growing tissues, however, respiration is seldom limited by substrate or enzyme capacity. Experiments with agents that uncouple respiratory metabolism from proton transfer, for example, reveal that normal rates are only a small fraction of capacity. In most tissues, the system is ultimately dependent upon the supply of ADP for ATP production. If ATP is not used rapidly in metabolism, respiration slows or stops. This "close coupling" is an important and useful concept. Shortage of ADP could slow production of intermediates, for example, in green tissues when chloroplasts take up ADP and Pi and release ATP to cytosol. More generally, it occurs when there is little use of ATP within the cells as may occur when growth is slow.

Genetic Control of Respiration–Biosynthesis
Like chloroplasts, mitochondria are thought to have resulted from ancient endosymbioses with bacteria and this is reflected in their circular, polyploid, DNA. The protein complexes of mitochondrial electron transport, like their counterparts in chloroplasts (Fig. 2) are composed of numerous subunits. Genetic control is much simpler than for photosynthesis, however. Glycolytic and TCA cycle enzymes are encoded by nuclear genes while mitochondrial DNA encodes most of the e- transfer complexes (Siedow and Umbach, 1995; Taiz and Zeiger, 1998). The alternative oxidase is under nuclear control. Isozymes are found for some glycolytic enzymes, the proportions of which may vary with tissue, developmental stage, and environment (Plaxton, 1996). This may be important for regulation and/or tissue-specific metabolism.

Because respiration is driven by demands for its products (ATP, reductant, C-skeletons), much of the genetic control resides with nuclear genes that determine form, structure, and composition (i.e., growth and development) of the whole plant.

A Functional Model of Respiration
Before 1940, when respiratory pathways and the role of ATP were not fully understood, research on respiration depended on measurement of substrate use or gas exchange. Microbiologists concerned with production efficiency of fermentation gave particular attention to the fate of substrate carbon. Duclaux (1898, cited by Pirt, 1965) was perhaps the first to distinguish between substrate used in microbial growth and that used in maintenance of cells. Animal scientists also recognized a maintenance component, in feed requirements and gas exchange, that included basal metabolism and was distinct from requirements for growth and work. Plant scientists, by contrast, focused in their early work on special cases where respiration rates were increased, for example, by ion uptake, ripening of fruit, or germination.

K.J. McCree (1970) demonstrated through regression analysis of gas exchange by white clover (Trifolium repens L.) plants of different sizes that whole-plant respiration could be partitioned into two components, one related to gross photosynthesis and the other to dry mass of living tissue. This opened a new era in plant research. Thornley (1970) showed that McCree's equation could be derived from Pirt's (1965) version of the Duclaux equation. Pirt's definition of maintenance as "energy consumed for functions other than production of new cell material" was carried forward by default. Penning de Vries (1975) identified protein turnover and active intracellular transport (e.g., maintenance of ion gradients) as the most important maintenance processes. Some workers distinguish intercellular transport as a third component, but this is normally distributed between growth and maintenance components.

Two useful definitions derive from Pirt's equations:

(1)
where {Delta}w is the amount of biomass formed, {Delta}SG is the amount of substrate used in growth and {Delta}SM is the amount of substrate used in maintenance during the same period of time, {Delta}t. {Delta}SM can be further defined as mW, where m is the "maintenance coefficient" and W is total living mass. Observed growth yield was recognized in the "growth efficiency" of Tanaka and Yamaguchi (1968). When mW is zero, or removed,

(2)

{Delta}SG is composed of two parts: {Delta}SR, which is completely respired to provide energy for synthesis (i.e., "growth respiration"), and {Delta}SB, which is retained in the new biomass as carbon skeletons. Total respiration for the period {Delta}t then is

(3)

The concept of close coupling of respiration with constructive use of ATP and reductant is implicit in these equations. Where it can be identified, however, a term for "wasteful respiration" can be included (Thornley, 1971).

Maintenance Respiration
Methods for estimating maintenance coefficients (m) in higher plants were reviewed earlier (Amthor, 1989). All are subject to some criticism. In particular, none has the ability to account for the high rates of protein turnover that occur in leaves in light and for interactions between respiration, photosynthesis, and photorespiration (see above and discussion by Amthor, 1994b). Values of m of 15 to 50 mg hexose g-1 biomass d-1 have been reported at temperatures near 25°C. The rate increases with temperature with Q10 of about 2 and with N content of the material. The increase in m with N content is expected on the basis that protein turnover is probably a major factor in maintenance. Zerihun et al. (1998) calculated turnover costs of 0.41 g glucose/g protein with 30% amino acid cycling. On this basis, protein turnover would account for only 3 to 20% of observed maintenance respiration. Maintenance also increases during senescence and McCree (1982) and others found that m tended to increase with growth rate. It is well to keep in mind, however, that rapid growth requires rapid respiration for biosynthesis. And, growth per unit respiration is greatest with efficient respiration.

Given the difficulties in measurement, few attempts have been made to determine the value of m under field conditions (all have employed McCree's regression method). Early work reviewed by Amthor (1989) and a more recent study by Mitchell et al. (1991) support the general conclusion that seasonal maintenance respiration about equals seasonal growth respiration. This question has also been approached theoretically with simulation models. For example, Ng and Loomis (1984) employed m values appropriate to each type of tissue in simulating total-season mW for a potato (Solanum tuberosum L.) crop; total maintenance equaled 21% of seasonal gross photosynthesis compared with 20% for growth respiration. The sensitivity of such models to values of m indicates that a small reduction in m might correspond to an appreciable advance in yield. Wilson (1982) made the promising discovery that ryegrass (Lolium perenne L.) selections having low rates of respiration in mature leaves had higher yields of simulated swards in growth chambers. The work was based on the common assumption that respiration of mature leaves is mainly maintenance although phloem loading and biosynthesis of amino acids and other compounds may occur there. Wilson's selections have been subjected to considerable study without resolving the basis for the smaller value of m and Kraus' (1992) report that the yield advantage disappeared when the simulated swards were grown at low density is a bit sobering. More recently, Earl and Tollenaar (1998) found a strong negative association across a series of maize hybrids between seasonal dry matter production and respiration rates of mature leaves.

It remains unclear whether crops already operate efficiently or whether reduction in m might be possible. This will require more knowledge about membrane leakage and turnover of individual proteins and about whether present rates of those processes are necessary for healthy plants. Much of our existing knowledge was obtained with plants grown in nutrient cultures under low PFD; good data for crops growing under field conditions (e.g., Mitchell et al., 1991) are rare and are needed.

Growth Respiration
Equation [2] offered a method for analysis of true growth yield that was elegantly pursued by Penning de Vries et al. (1974). By tracing "least-cost" biochemical pathways, they constructed balance sheets for use of substrate, ATP, and reductant and release of CO2 during synthesis of individual compounds. From these, they calculated general YG values (which they termed production values, PV) for carbohydrates, proteins, lipids, lignin, and other constituents of plant mass. Given a proximal analysis of biomass, this allows calculation of the amount of substrate and respiration involved in its synthesis if all goes well in the plant and least-cost pathways are actually used.

Penning de Vries et al. (1974) assumed a yield of 38 ATP per glucose oxidized (P:O = 3) (rather than about 30 as given above) causing a 3% overestimate of YG for maize in their Table 6. They showed in a sensitivity analysis (their Fig. 6) that variation of P:O between 2 and 3 had only a slight effect on YG because the bulk of carbon used in synthesis was retained in the C-skeleton of the product and use of reductant was as important as was use of ATP. YG values (Penning de Vries et al., 1983), in grams of product per gram of glucose consumed, for typical lipid (0.33), protein (0.40 from NO3-N and 0.62 with amide-N), and lignin (0.47) are much smaller than for carbohydrates (0.83) and organic acids (>1). Calculating protein yields starting with amide-N mimics metabolism during seed formation that utilizes N mobilized from vegetative tissues.

Comparison of YGs for various grains and types of biomass helps rationalize differences in yields among crops (Sinclair and de Wit, 1975; Penning de Vries et al., 1983). With other constituents of a seed held constant, YG declines almost 0.50% for each percentage increase in lipid concentration (replacing carbohydrate) and about 0.25% for each percentage increase in protein concentration (again, replacing carbohydrate). This illustrates why, for some purposes, RUE comparisons should be made in glucose or energy content equivalents. Comparing maize and soybean [Glycine max (L.) Merr.] grains starting with nitrate:

Maize
YG = 0.74 g grain g-1 glucose

Soybean
YG = 0.50 g grain g-1 glucose.

Thus, for the same amount of assimilate, soybean can yield only 2/3 as much as maize (soybean yield is even less when it depends on N2 fixation). Similar exercises explain why high-lysine grains, which have a smaller YG than normal grains, generally yield less (Mitra et al., 1979), and why suggestions for increasing proline and betaine contents as osmotica in moisture-stressed plants would be counter-productive.

YG can also be estimated from complete (McDermitt and Loomis, 1981) or partial (Vertregt and Penning de Vries, 1987; Williams et al., 1987) elemental analyses of biomass. Complete elemental analysis provides the reduction level of the biomass relative to standard states and closely correlates with heat of combustion of the biomass (McDermitt and Loomis, 1981; Jenkins and Ebeling, 1985). In line with the P:O sensitivity analysis of Penning de Vries et al. (1974), McDermitt and Loomis (1981) concluded that only about 20% of the energy content of glucose is lost in least-cost biosynthesis.

Growth respiration may be calculated directly with the pathway method of Penning de Vries et al. (1974). In their Table 6, they found that growth respiration (SR) for young maize plants with a large (23%) content of N compounds equaled 12 mmol CH2O g-1 biomass, or about 24% of substrate used for growth. Our Table 2, with about 7% N compounds, qr of 16, and RUE of 2.3 g MJ-1 intercepted begins with 129 mmol C in gross photosynthesis. Of this, 25 mmol C (19%) was expended in maintenance, 82 mmol C (64%) was retained in biomass, and 22 mmol went to growth respiration. Total respiration consumed 47 mmol C (36%). Thus pathway and elemental analyses provide useful hypotheses about how the productivity of crops varies with composition. The calculated values are in line with limited and somewhat imprecise experimental observations for good crops (Amthor, 1989) but we don't yet know the extent that crops attain those levels of efficiency.


    Assessments
 TOP
 ABSTRACT
 INTRODUCTION
 Photosynthetic Production
 Respiration and Biosynthesis
 Assessments
 REFERENCES
 
We have emphasized selected physiological aspects of photosynthesis and respiration to demonstrate how much remains to be accomplished in sorting out the regulation and function of various components. If there are inefficiencies in these systems and if the causes can be identified, they would represent legitimate targets for genetic manipulation. Properties of rubisco, alternative oxidase, and the photorespiration process already loom as opportunities for genetic manipulation. Those factors evolved and have survived during millions of years, however, indicating utilities of which we are as yet uncertain. With help of antisense RNAs and other methods, progress in deciphering genetic control is good, but much remains to be learned.

Serious reinvestigation of foliage canopies offers promise for important gains in photosynthetic productivity of crops. The first cycle of such research, begun over 50 yr ago, demonstrated the importance of such things as rapid and complete canopy cover, and the strong advantages of erect leaves in dense canopies and minimum interception by emergent reproductive structures. Those properties are now credited with contributing to yield progress in maize (Fischer and Evans, 1999). Our discussion of acclimation and protection processes was aimed at emphasizing fine details of canopy affairs about which we have information for only a few crops. Many more crops need to be studied in the way that Evans (1995) looked at alfalfa. Given the multifaceted, time-dependent nature of those issues, experimental work will need support from sophisticated simulation models to explore weaknesses in performance of present crops. To do that, the models will need to be constructed with fine details of morphology and physiology (Amthor and Loomis, 1996)—details that embrace such matters as variations in leaf and chloroplast structure and rubisco kinetics.

Understanding of photosynthetic and respiratory metabolism is now advanced enough to calculate potential efficiencies of using sunlight to produce plant biomass. These can be compared with best available observed efficiencies in order to obtain a hint of whether crops can be improved, and by how much, with respect to basic metabolic pathways and biophysical reactions. Our earlier analysis of wheat RUE (Loomis and Amthor, 1996), and our present one for maize, indicate that scope exists for improvement in yield (potential) for many crops. This scope exists within the context of existing intercepted (or absorbed) PAR. If radiation absorption can be further improved by more rapid canopy development or longer growing seasons, then potential yield will be improved further.


    ACKNOWLEDGMENTS
 
Our thanks to R.W. Pearcy for useful discussions of photosynthesis.

Received for publication December 12, 1998.


    REFERENCES
 TOP
 ABSTRACT
 INTRODUCTION
 Photosynthetic Production
 Respiration and Biosynthesis
 Assessments
 REFERENCES
 




This article has been cited by other articles: