The Physiological Ecology of C3-C4 Intermediate Eudicots in Warm Environments

THE PHYSIOLOGICAL ECOLOGY OF C3-C4 INTERMEDIATE

EUDICOTS IN WARM ENVIRONMENTS

Patrick John Vogan

A thesis submitted in conformity with the requirements for the degree of Doctorate of

Philosophy

Department of Ecology and Evolutionary Biology

University of Toronto

2010 ii

The Physiological Ecology of C3-C4 Intermediate Eudicots in Warm Environments

Patrick John Vogan, Doctor of Philosophy, 2010

Department of Ecology and Evolutionary Biology, University of Toronto

Abstract

The C3 photosynthetic pathway uses light energy to reduce CO2 to carbohydrates and other organic compounds and is a central component of biological metabolism. In C3 photosynthesis, CO2 assimilation is catalyzed by ribulose-1,5-bisphosphate carboxylase/oxygenase (Rubisco), which reacts with both CO2 and O2. While competitive inhibition of CO2 assimilation by oxygen is suppressed at high CO2 concentrations, O2 inhibition is substantial when CO2 concentration is low and O2 concentration is high; this inhibition is amplified by high temperature and aridity (Sage 2004). Atmospheric CO2 concentration dropped below saturating levels 25-30 million years ago (Tipple & Pagani

2007), and the C4 photosynthetic pathway is hypothesized to have first evolved in warm, low latitude environments around this time (Christin et al. 2008a). The primary feature of C4 photosynthesis is suppression of O2 inhibition through concentration of CO2 around Rubisco.

This pathway is estimated to have evolved almost 50 times across 19 angiosperm families

(Muhaidat et al. 2007), a remarkable example of evolutionary convergence. In several C4 lineages, there are species with photosynthetic traits that are intermediate between the C3 and

C4 states, known as C3-C4 intermediates. In two eudicot genera, Flaveria (Asteraceae) and

Alternanthera (Amaranthaceae), there is evidence that these species represented an intermediate state in the evolution of the C4 pathway (McKown et al. 2005; Sanchez-del Pino

2009). The purpose of this thesis is to ascertain the specific benefits to plant carbon balance

ii iii

and resource-use efficiencies of the C3-C4 pathway relative to C3 species, particularly at low

CO2 concentrations and high temperatures, factors which are thought to have been important in selecting for C3-C4 traits (Ehleringer et al. 1991). This will provide information on the particular advantages of the C3-C4 pathway in warm, often arid environments and how these advantages may have been important in advancing the initial stages of C4 evolution in eudicots. This thesis addresses the physiological intermediacy of previously uncharacterized

C3-C4 species of Heliotropium (Boraginaceae); the water- and nitrogen-use efficiencies of

C3-C4 species of Flaveria; and the photosynthetic performance and acclimation of C3, C4 and

C3-C4 species of Heliotropium, Flaveria and Alternanthera grown at low and current ambient CO2 levels and high temperature.

iii iv

Acknowledgements

I would like to recognize those who contributed to the completion of this thesis. Firstly, my graduate advisor, Dr. Rowan Sage for educating me in plant physiological ecology and spending countless hours reviewing and overseeing my research. Also, I would like to thank my parents, John and Donna Vogan, for their selfless support of my education and personal development for all these years, as well as the rest of my family. Finally, I must recognize the contributions of the innumerable friends and colleagues that I have met in the past five and a half years in Canada, the United States and the United Kingdom. There are far too many to list individually, but I want to express my sincerest gratitude to the graduate students, faculty members, technicians, horticulturists, post-doctoral fellows, administrative staff, research collaborators and friends from all walks of life who supported me in countless ways during the completion of my degree.

iv v

Table of Contents

Abstract……………………………………………………………………………………….ii

Acknowledgements………………………………………………………………...….…...... iv

Table of Contents...... v

List of Tables...... viii

List of Figures...... ix

Chapter One – Introduction...... 1

I.) C3 photosynthesis and the problem of photorespiration...... 1

i.) C3 photosynthesis: mechanism and origins...... 1

ii.) The photorespiratory cycle...... 2

iii.) The effects of photorespiration on CO2 assimilation...... 4

iv.) Carbon balance and fitness of C3 plants under low CO2 conditions...... 6

II.) Atmospheric CO2 and the rise of C4 photosynthesis...... 9

i.) The mechanism of C4 photosynthesis...... 9

ii.) The physiological ecology of C4 photosynthesis...... 16

iii.) The evolutionary origins of C4 photosynthesis...... 19

III.) C3-C4 intermediate photosynthesis...... 23

i.) The anatomy and physiology of C3-C4 photosynthetic plants...... 23

ii.) The taxonomic and ecological distributions of C3-C4 photosynthesis...... 26

IV.) Models of C4 evolution...... 28

V.) Thesis objectives...... 33

Chapter Two – The functional significance of C3-C4 intermediate traits in Heliotropium

(Boraginaceae): gas exchange perspectives...... 36

v vi

Abstract...... 36

Introduction...... 36

Materials and Methods...... 39

Results...... 43

Discussion...... 55

Conclusion...... 59

Chapter Three – Water-use and nitrogen-use efficiency of C3-C4 intermediate species of

Flaveria Juss. (Asteraceae)...... 61

Abstract...... 61

Introduction...... 61

Materials and Methods...... 66

Results...... 69

Discussion...... 90

Conclusion...... 96

Chapter Four - Photosynthetic performance and acclimation of C3, C4 and C3-C4 species from three eudicot genera grown at low CO2 concentrations...... 97

Abstract...... 97

Introduction...... 98

Materials and Methods...... 103

Results...... 105

Discussion...... 133

Conclusion...... 139

Chapter Five – Discussion...... 141

vi vii

I.) C3-C4 intermediacy in Heliotropium and the development of new model systems to

study C4 evolution...... 142

II.) Water- and nitrogen-use efficiencies of C3-C4 Flaveria species...... 145

III.) The effects of low CO2 on carbon balance and photosynthetic acclimation in C3,

C4 and C3-C4 species from multiple evolutionary lineages...... 150

References...... 153

vii viii

List of Tables

Table 2.1 – Habitat description and collection information for Heliotropium plants in gas exchange study……………………………………………………………………………....46

Table 2.2 – Carbon isotope discrimination and gas exchange parameters for Heliotropium species, organized by species……………………………………………………………..…47

Table 2.3 – Gas exchange parameters in Heliotropium, organized by photosynthetic type…………………………………………………………………………..………………49

Table 2.4 – Regression data for response of CO2 compensation point to O2 concentration for Heliotropium species………………………………………………………………………...50

Table 2.5 – Stomatal conductances of Heliotropium species measured at four cuvette CO2 concentrations………………………………………………………………………………..51

14 Table 3.1 – CO2 compensation points and fixation of C to C-4 compounds in Flaveria species………………………………………………………………………………………..73

Table 3.2 – Instantaneous water-use efficiency and linear regression data for response of net CO2 assimilation rate to stomatal conductance in Flaveria species...... 74

Table 3.3 – Instantaneous nitrogen-use efficiency and quadratic regression data for response of net CO2 assimilation rate to leaf nitrogen content in Flaveria species…………………...75

Table 3.4 – Linear regression data for response of net CO2 assimilation rate to leaf Rubisco content...... 76

Table 4.1 – Gas exchange parameters measured at 30°C and 40°C for Alternanthera, Flaveria and Heliotropium species of three different photosynthetic types……………….110

Table 4.2 – Data for response of net CO2 assimilation rate to temperature for Alternanthera, Flaveria and Heliotropium species grown at two CO2 concentrations…………………….112

Table 4.3 – Leaf tissue chemistry and nitrogen allocation patterns for Alternanthera, Flaveria and Heliotropium species grown at two CO2 concentrations………………………………113

Table 4.4 – Photosynthetic water- and nitrogen-use efficiencies, stomatal conductances and ratios of intercellular-to-ambient CO2 concentration for Alternanthera, Flaveria and Heliotorpium species measured at two cuvette CO2 concentrations……………………….115

viii ix

List of Figures

Figure 1.1 – Ratio of photorespiration to photosynthesis in response to temperature and intercellular CO2 concentration in C3 leaves, from Ehleringer et al. 1991…………………....5

Figure 1.2 – Diagram of photosynthetic metabolism in C4 NADP-ME subtype…………….11

Figure 1.3 – Diagram of photosynthetic metabolism in C4 NAD-ME subtype……………...12

Figure 1.4 – Diagram of photosynthetic metabolism in C4 PCK subtype…………………...13

Figure 2.1 – Response of net CO2 assimilation rate to intercellular CO2 concentration in Heliotropium species………………………………………………………………………...52

Figure 2.2 – Response of CO2 compensation point to O2 concentration in Heliotropium species………………………………………………………………………………………..54

Figure 3.1 – Response of net CO2 assimilation rate to stomatal conductance in Flaveria species, with linear regressions………………………………………………………………77

Figure 3.2 – Response of net CO2 assimilation rate to leaf nitrogen content in Flaveria species, with quadratic regressions…………………………………………………………..79

Figure 3.3 – Response of net CO2 assimilation rate to leaf Rubisco content in Flaveria species, with linear regressions………………………………………………………………81

Figure 3.4 – Intrinsic water-use efficiency and intercellular CO2 concentration of Flaveria species in response to fixation of 14C to C-4 compounds……………………………………83

Figure 3.5 – Rubisco-use efficiency and leaf Rubisco content of Flaveria species in response to fixation of 14C to C-4 compounds………………………………………………………...84

Figure 3.6 – Nitrogen allocation to Rubisco and thylakoid components in Flaveria species in response to total leaf nitrogen content……………………………………………………….86

Figure 3.7 – Response of net CO2 assimilation rate to intercellular CO2 concentration for three Flaveria species of different photosynthetic types, including supply function………..88

Figure 3.8 – Phylogeny of Flaveria, from McKown et al. 2005…………………………….89

Figure 4.1 – Response of net CO2 assimilation rate to intercellular CO2 concentration for three Alternanthera species grown at two CO2 concentrations and measured at leaf temperatures of 30°C and 40°C…………………………………………………………….116

ix x

Figure 4.2 – Response of net CO2 assimilation rate to intercellular CO2 concentration for three Flaveria species grown at two CO2 concentrations and measured at leaf temperatures of 30°C and 40°C……………………………………………………………………………...117

Figure 4.3 – Response of net CO2 assimilation rate to intercellular CO2 concentration for three Heliotropium species grown at two CO2 concentrations and measured at leaf temperatures of 30°C and 40°C…………………………………………………………….118

Figure 4.4 – Response of net CO2 assimilation rate to leaf temperature for three Alternanthera species grown at two CO2 concentrations………………………………...... 119

Figure 4.5 – Response of net CO2 assimilation rate to leaf temperature for three Flaveria species grown at two CO2 concentrations………………………………………………….121

Figure 4.6 – Response of net CO2 assimilation rate to leaf temperature for three Heliotropium species grown at two CO2 concentrations………………………………………………….123

Figure 4.7 – Response of stomatal conductance to intercellular CO2 concentration for three Alternanthera species grown at two CO2 concentrations and measured at leaf temperatures of 30°C and 40°C……………………………………………………………………………...125

Figure 4.8 – Response of stomatal conductance to intercellular CO2 concentration for three Flaveria species grown at two CO2 concentrations and measured at leaf temperatures of 30°C and 40°C……………………………………………………………………………...126

Figure 4.9 – Response of stomatal conductance to intercellular CO2 concentration for three Heliotropium species grown at two CO2 concentrations and measured at leaf temperatures of 30°C and 40°C……………………………………………………………………………...127

Figure 4.10 – Response of the ratio of intercellular-to-ambient CO2 concentration to ambient CO2 concentration for three Alternanthera species grown at two CO2 concentrations and measured at leaf temperatures of 30°C and 40°C…………………………………………..128

Figure 4.11 – Response of the ratio of intercellular-to-ambient CO2 concentration to ambient CO2 concentration for three Flaveria species grown at two CO2 concentrations and measured at leaf temperatures of 30°C and 40°C……………………………………………………..129

Figure 4.12 – Response of the ratio of intercellular-to-ambient CO2 concentration to ambient CO2 concentration for three Heliotropium species grown at two CO2 concentrations and measured at leaf temperatures of 30°C and 40°C…………………………………………..130

Figure 4.13 – Relative stomatal limitation for species of Alternanthera, Flaveria and Heliotorpium grown at two CO2 concentrations…………………………………………...131

Figure 4.14 – Leaf temperatures of Heliotropium convolvulaceum measured over four days in August and September, 2006 near Overton, NV, USA………………………………….132

x Chapter One – Introduction

I) C3 photosynthesis and the problem of photorespiration

Ii.) C3 photosynthesis: mechanism and origins

Carbon dioxide assimilation has been a central component of biological metabolism since the beginning of life on Earth approximately 3.5 billion years ago, and the assimilation of carbon via the Calvin Cycle represents the primary way in which inorganic carbon enters the biosphere (Tabita et al. 2007, 2008). Energy from sunlight is used to reduce carbon dioxide, thus producing a means for organic compounds to form and for photosynthetic organisms to synthesize energy-containing molecules for the purposes of growth and reproduction.

Oxygenic photosynthesis, that is, photosynthesis in which H2O is the primary electron donor and is oxidized to O2, likely originated at least 2.9, and perhaps as many as

3.5 billion years ago (Nisbet et al. 2007). It is thought that this was preceded by some form of anoxygenic photosynthesis (Olson & Blankenship 2004), and, in any case, evidence from

13 12 C/ C ratios indicates that some form of enzymatic CO2 fixation has existed for about 3.5 billion years (Hayes 1994). The enzyme that has mostly been responsible for CO2 fixation is ribulose 1,5-bisphosphate carboxylase/oxygenase (Rubisco), which has been implicated in all forms of oxygenic photosynthesis known to have existed (Nisbet et al. 2007).

Rubisco catalyzes the fixation of one molecule of CO2 to the pentose-bisphosphate sugar ribulose-1,5-bisphosphate (RuBP), yielding two molecules of the three-carbon phosphoglyceric acid (PGA). Phosphoglyceric acid is then reduced to glyceraldehyde-3- phosphate by ATP and NADPH produced in the light reactions. While five-

1 2 sixths of the PGA produced by Rubisco is recycled to regenerate RuBP in the Calvin Cycle, the remainder is used to synthesize carbohydrates and to provide carbon skeletons for fundamental cellular components such as nucleic acids and proteins (Leegood et al. 2000).

Iii.) The photorespiratory cycle

In addition to its function as a carboxylase, Rubisco also catalyzes the fixation of one molecule of O2 to RuBP, yielding one molecule of PGA and one molecule of the two-carbon phosphoglycolate (PG). In contrast to RuBP carboxylation, RuBP oxygenation is regarded as a wasteful side reaction of Rubisco (Andrews & Lorimer 1987). While RuBP oxygenation has been shown to partially alleviate photoinhibition by providing an alternative pathway for energetic electrons (Osmond 1981; Wingler et al. 2000; Bai et al. 2008), this enzymatic function is an inherent consequence of the active site chemistry of Rubisco and its reactions with CO2 rather than an adaptive response to photoinhibition (Lorimer & Andrews 1981;

Andrews & Lorimer 1987; Ehleringer et al. 1991). Underscoring this point, RuBP oxygenation can even be catalyzed by Rubisco from anaerobic organisms (Andrews &

Lorimer 1987). The oxygenase activity of Rubisco would not have been significant for the first one to two billion years of the enzyme’s history because atmospheric O2 concentrations were low and CO2 was high (Lowe 1994), but atmospheric O2, and the potential for RuBP oxygenation, increased greatly after this time.

In the initial reaction of the photorespiratory pathway, RuBP is oxygenated to produce one molecule of PGA, which can be metabolized in the Calvin Cycle, and one molecule of PG. Phosphoglycolate has no apparent useful metabolic function in the plant cell, and its accumulation is toxic to plants. Arabidopsis plants deficient in the pathway to

metabolize PG exhibit a nearly 100% reduction in net CO2 assimilation rate under current ambient CO2 and O2 concentrations (Somerville & Ogren 1979), produce lesions on their leaves, and shortly thereafter die (Somerville & Ogren 1982; Ogren 1984). To prevent the accumulation of PG, plants have evolved a 15-step pathway involving three organelles

(chloroplast, peroxisome, and mitochondrion) to metabolize PG. In the photorespiratory cycle, two molecules of PG react to produce one molecule of PGA and one CO2 at a net cost of two ATP and two reducing equivalents in the form of NADPH (Sage 1999; Douce &

Heldt 2000). This energy investment and the loss of previously-fixed carbon involved in the metabolism of PG, as well as the competitive inhibition of RuBP carboxylation by O2, reduces whole plant carbon balance (Sharkey 1988).

In order for oxygenase activity to become significant, O2 concentration in solution must be about 10-fold greater than CO2 concentration (Jordan & Ogren 1984). Because of the different solubilities of the two molecules, atmospheric concentration of O2 must be about 100 times greater than CO2 concentration in order to have 10-fold greater concentration of O2 in solution at 30°C (von Caemmerer & Quick 2000). Prior to the

Carboniferous period (350-300 Ma), atmospheric CO2 levels were above 1,000 µmol CO2 mol-1 air (hereafter µmol mol-1), the level needed to effectively saturate Rubisco and suppress RuBP oxygenation at 20% oxygen, and they declined from around 800,000 µmol mol-1 at 3.5 billion years ago to around 2,000 µmol mol-1 at the beginning of the

Carboniferous (Sage 2004). During the Carboniferous, CO2 concentrations dropped below

-1 500 µmol mol , but a subsequent rise in atmospheric CO2 concentration to near 2,000 µmol mol-1 mostly suppressed oxygenase activity again from 300-35 Ma (Sage 1999; Sage 2004).

During the Cenozoic Era (65 Ma to present), atmospheric O2 concentrations have been above

20% of atmospheric pressure while CO2 concentration has dropped from about 2,500 µmol mol-1 to below 300 µmol mol-1 (Berner 1994; Tipple & Pagani 2007). Beginning around 35

Ma, CO2 concentrations fell below the level needed to effectively saturate Rubisco and suppress RuBP oxygenation, and the potential for photorespiration increased substantially after this time.

Iiii.) The effects of photorespiration on CO2 assimilation

Photorespiration rate increases with temperature and is inversely related to CO2 concentration (Ehleringer et al. 1991). The effect of temperature is partly a result of the different solubilities of the two gases; CO2 concentration in solution declines more rapidly with rising temperature than O2 concentration (Ku & Edwards 1977; Jordan & Ogren 1984).

Also, Rubisco specificity for CO2 as a substrate relative to O2 decreases as temperature increases (Jordan & Ogren 1984). The net result is a non-linear increase in the proportion of photorespiration to photosynthesis with increasing temperature and decreasing CO2 (Fig.

1.1). At 30°C, photorespiration can reduce net CO2 assimilation rates by over 30% at CO2 levels of 350 µmol mol-1 (Sharkey 1988).

The negative effect of photorespiration on leaf-level photosynthesis is well-illustrated by the effect of varying O2 concentration on the response of net CO2 assimilation rate (Anet) to intercellular CO2 concentration in the leaf (Ci). In tomato, net photosynthetic rate at 30°C is enhanced by over 25% when O2 concentration is changed from 21% to 2% (Ku et al.

1983). The CO2 compensation point (Γ), where CO2 influx from photosynthesis is equal to

CO2 efflux from photorespiration and mitochondrial respiration, decreases from 54 to 9 µmol

-1 mol . Carboxylation efficiency (CE), the amount of increase in Anet per unit of CO2 added in

Figure 1.1. Ratio of photorespiration to photosynthesis in response to intercellular CO2 concentration and leaf temperature. Arrows indicate typical C3 leaf intercellular CO2 concentration under current atmospheric CO2 concentrations. From Ehleringer et al. 1991.

the initial, linear portion of the A/Ci response, increases by 212%. Similar enhancements in

Anet and CE and similar decreases in Γ under low O2 have been observed in soybean

(Holaday et al. 1984) as well as C3 species of Flaveria (Asteraceae) (Ku et al. 1991) and

Alternanthera (Amaranthaceae) (Rajendrudu et al. 1986).

The effects of photorespiration on leaf-level photosynthesis are also manifested in the effect of varying CO2 concentration on the response of Anet to leaf temperature. CO2 addition has a strong effect on the thermal optimum of photosynthesis, pushing it to higher values by suppressing photorespiration. For example, in the C3 plant Nerium oleander grown at daytime temperatures of 45°C, the thermal optimum of C3 photosynthesis shifts from 37 to

-1 45°C when CO2 concentration is raised from 330 to 850 µmol mol , and net CO2 assimilation rate increases by over 100% at the thermal optimum (Berry & Bjorkman 1980).

In sweet potato grown at daytime temperatures of 27°C, the thermal optimum shifts from near 22°C at 140 µmol mol-1, to about 26°C at 250 µmol mol-1, and to about 31°C at 500

µmol mol-1 (Cen & Sage 2005). Photosynthetic rates in sweet potato increase by approximately 100% at the thermal optimum as well at each of these stages. Together, these results demonstrate the considerable negative impact of RuBP oxygenation on the leaf’s ability to assimilate CO2 efficiently at ambient and sub-ambient CO2 concentrations, particularly at high temperatures.

Iiv.) Carbon balance and fecundity of C3 plants under low CO2 conditions

Photorespiration rate is increased at low intercellular CO2 concentrations and high temperatures. One way in which intercellular CO2 concentration is reduced is when ambient

CO2 concentration is low, and the effects of low ambient CO2 concentrations on

photosynthesis, growth and reproduction in C3 plants have been studied extensively in recent years. These experiments demonstrate a substantial interaction between temperature and ambient CO2 concentration and a strongly negative impact on plant carbon balance and fecundity in a laboratory setting.

Several studies have examined the effects of CO2 reduction on C3 plants. For example, total biomass is reduced in Abutilon theophrasti by 24% and 92% after 35 days

-1 growth at 270 and 150 µmol mol , respectively, compared to ambient CO2 concentrations of

350 µmol mol-1 (Dippery et al. 1995). Similar reductions in Abutilon biomass were also observed by Ward et al. (1999) under the same CO2 treatments. In wheat, oat and wild mustard grown near 200 µmol mol-1, total biomass is 50% lower than in plants grown at 340

µmol mol-1 after 15 weeks (Polley et al. 1992, 1993). Campbell et al. (2005) observed a nearly 50% reduction in biomass in tobacco plants grown at 190 µmol mol-1 relative to ambient at 19/15°C and 30/25°C, and also determined that plants grown at or below 150

µmol mol-1 remained in the seedling stage (total leaf area <1 cm2) over three weeks. As seedlings, these plants would have been considerably more vulnerable to stochastic fluctuations in temperature and water availability during this period than plants grown at higher CO2 concentrations, and, consistently, plant mortality increased by 100-400% in

-1 treatments below 270 µmol mol (Campbell et al. 2005). The effects of reduced CO2 on C3 plant growth are amplified at high temperatures. Cowling & Sage (1998) found that

Phaseolus plants grown at 380 µmol mol-1 exhibited a 22% reduction in biomass and 42% reduction in leaf area when grown at 36/29°C relative to 25/20°C. At 200 µmol mol-1, total biomass was reduced by almost 70% at 36/29°C relative to 25/20°C. In tobacco grown at

190 and 270 µmol mol-1, biomass declined by about 50% for plants grown at 30/25°C compared to the 19/15°C treatment (Campbell et al. 2005).

These reductions in biomass and growth rate are generally accompanied by reductions in reproductive output as well. Campbell et al. (2005) found that total fruit and flower number, seed number, seed weight, and offspring germination rates in tobacco were significantly reduced by 14-63% from 150 to 100 µmol mol-1. Ward et al. (2000) determined that seed set in Arabidopsis was significantly reduced at 200 µmol mol-1, and Dippery et al.

(1995) observed abortion of all flowers and buds before anthesis in Abutilon grown at 150

µmol mol-1. These reductions in biomass and seed set demonstrate the pronounced effects of low CO2 concentrations on the fitness of C3 plants.

An additional factor influencing C3 plant performance under high temperature and low CO2 conditions is aridity. In natural environments, low intercellular CO2 can occur when atmospheric CO2 concentration is low, but it can also occur when stomatal conductance is low, as often happens in arid or saline environments where the imperative of water conservation results in low stomatal conductance and, consequently, reduced rates of

CO2 diffusion into the leaf (Sage & Reid 1995; Sage & Pearcy 2000) As such, aridity is a substantial constraint on the plant’s ability to increase stomatal conductance, and thus Ci, as a means of reducing photorespiration, further exacerbating the negative effects of photorespiration in warm environments.

These studies demonstrate the substantial constraints on the productivity and reproductive output of C3 plants in low CO2 environments, and the further impacts of high temperature and aridity. When such conditions have occurred on Earth, a considerable

selective pressure likely existed for C3 plants to improve carbon economy and to increase efficiency of CO2 assimilation (Sage & Coleman 2001; Cowling 2001).

II.) Atmospheric CO2 and the rise of C4 photosynthesis

IIi.) The mechanism of C4 photosynthesis

-1 Atmospheric CO2 has fluctuated between 180 and 300 µmol mol for the past 35 Ma, not including anthropogenic CO2 emissions since the Industrial Revolution (Pagani et al.

1999; Zachos et al. 2001). Prior to that, atmospheric CO2 levels were greater than 1,000

-1 µmol mol , and this high level of CO2 would mostly have suppressed the oxygenase function of Rubisco. However, as atmospheric CO2 concentration dropped during the mid-Cenozoic,

C3 plants would have experienced much higher rates of photorespiration and greatly reduced rates of CO2 assimilation, growth and reproduction. During the glacial periods of the

Pleistocene epoch (1.8-0.01 Ma), for example, CO2 levels ranged between 180 and 220 µmol mol-1 (Sage & Coleman 2001). Stable isotope measurements of Juniperus fossils from this

-1 period indicate a Ci of ~113 µmol mol (Ward et al. 2005). In such low CO2 environments, and especially in areas marked by high temperature and aridity, strong evolutionary pressure likely existed for ways to make CO2 assimilation more efficient.

The C4 photosynthetic pathway, which is estimated to have first appeared 25-32 Ma

(Kellogg 1999; Cerling 1999; Christin et al. 2008a), successfully suppresses photorespiration and increases net CO2 assimilation rate at low CO2 and high temperature by isolating

Rubisco within the leaf and concentrating CO2 around it. While the pathway has arisen almost 50 times independently (Muhaidat et al. 2007), the common feature in all these lineages is the spatial separation of initial carbon assimilation from carbon reduction into

different parts of the leaf (or in the case of single-celled C4 species, into different parts of the cell).

In the leaves of most C4 plants, Rubisco is localized to the cells around the veins, commonly termed “bundle-sheath cells,” which are substantially larger in C4 than in C3 leaves and contain a significantly higher proportion of chloroplasts and mitochondria

(Dengler & Nelson 1999). This anatomical arrangement is termed “Kranz” anatomy. Initial

CO2 fixation is not catalyzed by Rubisco, but is instead carried out in the mesophyll by phosphoenolpyruvate carboxylase (PEPC). Because PEPC does not react with oxygen

(Edwards et al. 2004), current atmospheric O2 concentrations do not inhibit CO2 assimilation by PEPC. The first reaction of the C4 pathway (reactions summarized in Figs. 1.2-1.4) is the

- hydration of CO2 to HCO3 by carbonic anhydrase. Then, PEPC catalyzes the reaction of

- HCO3 with phosphoenolpyruvate (PEP) to yield one molecule of the four-carbon oxaloacetate (OAA), from which the C4 pathway derives its name. Oxaloacetate is subsequently converted to either malate or aspartate, depending upon the species. These four-carbon compounds then diffuse into the bundle-sheath cells via plamsodesmata, and there are three distinct C4 biochemical subtypes that employ slightly different mechanisms to release CO2 in bundle-sheath cells. In species using NADP-malic enzyme (NADP-ME), malate is decarboxylated to yield one CO2 and one pyruvate. In species using NAD-malic enzyme (NAD-ME), aspartate diffuses into the bundle-sheath where it is converted to OAA by aspartate aminotransferase; OAA is subsequently converted to malate by NAD malate dehydrogenase, and malate is then decarboxylated by NAD-ME to yield one CO2 and one pyruvate. In species using PEP carboxykinase (PCK), aspartate and malate both diffuse into bundle-sheath cells. Malate can be decarboxylated by NAD-ME, and aspartate is converted

NADP-ME subtype

Mesophyll cell Bundle-sheath cell

Pyruvate

ATP Rubisco Pyruvate CO2 PPDK RuBP AMP + PGA 2 Pi Malate NADP-ME Calvin Cycle PEP PEPC OAA HCO - Triose-P CO2 3 CA

Figure 1.2. Diagram of C4 metabolism in NADP-ME subtype. Enzymes are underlined.

NAD-ME subtype

Mesophyll cell Bundle-sheath cell

Pyruvate Rubisco Pyruvate CO2 RuBP ATP PPDK Calvin PGA AMP NAD-ME Aspartate Cycle + 2 Pi Malate Malate PEP PEPC OAA Triose-P HCO - CO2 CA 3

Figure 1.3. Diagram of C4 metabolism in NAD-ME subtype. Enzymes are underlined.

PCK subtype

Mesophyll cell Bundle-sheath cell

PPDK Pyruvate Rubisco ATP NAD-ME CO2 RuBP AMP Malate + 2 Pi CO2 PEP Calvin PGA PEPC Cycle PEP OAA

- HCO3 PCK CA Aspartate OAA CO2 Triose-P

Figure 1.4. Diagram of C4 metabolism in PCK subtype. Enzymes are underlined.

to OAA by aspartate aminotransferase which is subsequently decarboxylated by PCK to yield one PEP and one CO2. PEP produced in OAA decarboxylation can diffuse back into the mesophyll to continue the C4 cycle, and pyruvate produced in malate decarboxylation is converted to PEP by pyruvate Pi dikinase (PPDK) in the mesophyll (Furbank et al. 2000), a process that requires 2 ATP per molecule of PEP produced. Overall, the C4 cycle in all three subtypes entails an energy requirement essentially equal to the 2 ATP consumed by PPDK

(Furbank et al. 2000).

The net result of C-4 acid decarboxylation in the bundle-sheath is an effective CO2 pump in which bundle-sheath CO2 concentration increases to levels far greater than ambient

(> 1,000 µmol mol-1), largely suppressing RuBP oxygenation and greatly increasing carboxylation efficiency of Rubisco, particularly at high temperatures. The bundle-sheath cell walls in many C4 species are modified to minimize CO2 diffusion out of these cells after

C-4 acid decarboxylation in order to maintain high concentrations in the symplast. For example, in many C4 monocots, the bundle-sheath cell wall is impregnated with high amounts of suberin (Dengler & Nelson 1999). Other species may concentrate their chloroplasts along the interior surface of the bundle-sheath cells to increase the path length of

CO2 diffusion or reduce bundle-sheath surface area exposed to intercellular air spaces

(Dengler & Nelson 1999; von Caemmerer & Furbank 2003; Sage 2004).

The C4 pump alters the gas exchange physiology of photosynthesis from C3 patterns in several characteristic ways. For one, the A/Ci response of C4 photosynthesis differs substantially from that of C3 species. The CO2 compensation point (Γ) is dramatically lower

-1 in C4 than in C3 species. At 30°C, Γ of C3 species is between 50-60 µmol mol while in C4

-1 species it is 0-5 µmol mol , meaning that the C4 pathway concentrates CO2 so

effectively that leaves can maintain positive rates of net CO2 assimilation with only minute concentrations of CO2 in the surrounding air. Another change is a dramatic increase in carboxylation efficiency, as measured by the initial, linear portion of the A/Ci response. This value is approximately 4-8 times greater in C4 leaves than in C3 leaves measured at or near

30°C (Ku et al. 1983; Sudderth et al. 2007; Vogan et al. 2007). The C4 pump ensures that a given increase in CO2 concentration in the mesophyll will result in a proportionately greater amount of CO2 assimilation in a C4 than in a C3 leaf. Also, because C4 photosynthesis concentrates CO2 around Rubisco to near-saturating values, the C4 photosynthetic apparatus

-1 is CO2-saturated near Ci values of 150-200 µmol mol , meaning that it operates close to its maximum rate under current ambient conditions and those of prior atmospheres of the past

35 Ma. By contrast, C3 leaves at or above 30°C will not saturate Rubisco until ambient CO2 concentrations approach 700-1,000 µmol mol-1 (von Caemmerer et al. 1997; Sage 2002).

Another feature of the CO2-concentrating mechanism is an increase in the thermal optimum of photosynthesis. When photorespiration is reduced in C3 leaves by changing O2 concentration from 21% to 2%, there is a substantial increase in the thermal optimum of photosynthesis as well as the rate of net CO2 assimilation at the thermal optimum. Similar results have been gleaned from studies of C4 species. The photosynthetic thermal optimum

-1 of the C4 eudicot Amaranthus retroflexus is 34°C at 360 µmol mol and 31°C at 180 µmol

-1 mol while the thermal optimum of the ecologically similar C3 species Chenopodium album is 31°C and 22°C under the same conditions, respectively (Sage 2002). In A. retroflexus, photosynthetic rate at the thermal optimum is 33% and 100% greater than in C. album at those CO2 concentrations, respectively. In Flaveria, C4 species exhibit thermal optima of 25-

-1 30°C at 325 µmol mol compared to the C3 range of 20-22°C (Ku et al. 1991). It is also of

note that the plant with the highest recorded photosynthetic thermal optimum is a C4 species,

Tidestromia oblongifolia (Amaranthaceae), which exhibits a value of 47°C (Bjorkman et al.

1972).

Together, the greater rates of net CO2 assimilation at low CO2 and high temperature provide marked advantages for C4 plants when grown in low CO2 experiments. Dippery et al. (1995) grew Abutilon theophrasti (C3) and Amaranthus retroflexus (C4) under four CO2 concentrations (700, 350, 270 and 150 µmol mol-1) at 28/22°C, and they observed a significant reduction in total biomass, growth rate and seed set in Ab. theophrasti under lower CO2 treatments. By contrast, Am. retroflexus exhibited no significant change in these parameters between any treatments. Ward et al. (1999) found a nearly identical result in total biomass and growth rate in these species as well (without measuring seed set). The ultimate outcome of these alterations in gas exchange physiology is that the C4 pathway enhances carbon balance and reproductive output at low CO2 and high temperature and results in greater growth rate, biomass production and fitness under these conditions.

IIii.) The physiological ecology of C4 photosynthesis

The greater efficiency and higher CO2 assimilation rates of C4 photosynthesis at warm temperatures have resulted in a marked ecological advantage for these species under warm, high light conditions. Because of the energetic investment required to operate the C4 pump, however, C4 species have an advantage over C3 species only at temperatures where energy investment in photorespiration exceeds the ATP requirement of the C4 pump. As a result, C3 species are favored at the lower temperatures characteristic of high latitudes and elevations, and the threshold of this interaction is in the range of 22-30°C (Sage & Kubien

2003). The result is a much greater frequency of C4 species at low than high latitudes (Sage et al. 1999). The global distribution of C4 species correlates strongly with latitude and growth-season temperature (Teeri & Stowe 1976; Sage et al. 1999; Wan & Sage 2001). C4 grasses comprise roughly 75-100% of grass species in most sub-tropical and tropical grassland ecosystems, while cool temperate grasslands tend to be dominated by C3 species, and the frequency of C4 monocots at latitudes above 60° latitude is close to nil (Teeri &

Stowe 1976; Sage et al. 1999). C4 eudicots follow a similar pattern of distribution according to latitude, though their distribution correlates more strongly to aridity than temperature

(Stowe & Teeri 1978), and they do not approach the frequency of C4 monocots. C4 eudicots can locally dominate in some arid systems in Australia (Atriplex shrubs), Central Asia

(chenopod shrubs and Calligonum spp.) and western North America (Atriplex shrubs) (Stowe

& Teeri 1978; Sage et al. 1999).

In addition to improved carbon balance at high temperatures and low CO2 concentrations, C4 species are marked by higher resource-use efficiencies than C3 species.

The water-use efficiency of photosynthesis (PWUE) is generally measured as the amount of

CO2 assimilated per unit of water transpired (A/E). Because the CO2-concentrating mechanism of C4 leaves saturates Rubisco at relatively low values of Ci, these species can maintain lower stomatal conductance (and thus conserve water) without experiencing a corresponding decline in assimilation rate to the degree that a C3 plant would (Farquhar &

Sharkey 1982). Stomatal conductance is about 50-70% lower in a C4 leaf at a given rate of net CO2 assimilation (Schulze & Hall 1982), and, as a result, PWUE is approximately twice as great in C4 as in C3 species above 25°C (Sage & Pearcy 1987; Sage 2001; Vogan et al.

2007). In general, this means that C4 plants in warm environments can assimilate more CO2

18 and have higher growth rates (both above and below ground) for a given amount of water used (Sage & Pearcy 2000). Increased PWUE also provides an advantage for C4 species in tolerating very arid and saline environments. In hot deserts of North America, Africa and

Australia, the C4 pathway may facilitate plant growth in areas where very high temperature and low precipitation may exclude C3 species altogether (Schulze et al. 1996; Sage & Pearcy

2000). C4 species in these extremely xeric environments can tolerate conditions that may either induce potentially lethal evaporative demands on C3 leaves or require extremely low stomatal conductances, which would result in substantially lower Ci and greater photorespiratory inhibition of photosynthesis (Sage 2001). In coastal salt marshes of eastern

North America, C4 species occupy habitats that are far north of their normal range limit in non-saline habitats. The increased PWUE of C4 species reduces the amount of water (and consequently, salt) absorbed to achieve a given photosynthetic rate. As such, C4 grasses such as Spartina and Distichilis are frequent in coastal salt marshes of eastern North America as far north as Atlantic Canada, a latitude where C4 species are infrequent in terrestrial environments (Archibold 1995).

In addition to increased water-use efficiency, C4 species also exhibit photosynthetic nitrogen-use efficiencies (PNUE) that are 2 to 4 times greater than C3 species above 25°C

(Brown 1978; Schmitt & Edwards 1981; Li 1993; Sage & Pearcy 1987; Sage 2001). The greater efficiency of Rubisco in C4 photosynthesis means that less of the enzyme is required in the leaf to assimilate a given amount of CO2, and correspondingly, leaf Rubisco levels are

2-4 times lower in C4 than in C3 leaves (Sage et al. 1987; Sage & Seemann 1993). Because

Rubisco is the largest individual N investment in the C3 leaf, comprising 15-30% of total leaf nitrogen (Evans 1989), this substantial reduction in leaf Rubisco content, accompanied by a

19 reduction in photorespiratory enzymes and thylakoid protein, results in much higher values of PNUE (Sage et al. 1987; Ku et al. 1991). While high soil N content obscures this potential ecological advantage of C4 species in many natural environments (Sage & Pearcy

2000; Ripley et al. 2008), there may be an advantage under low N conditions. In central

North American prairies containing grass species of both photosynthetic types, C4 species exhibited greater biomass production when N fertilization was withheld, and this difference disappeared when fertilizers were applied, as C3 species completely crowded out C4 competitors (Wedin & Tilman 1996). C4 grasses also absorbed relatively more soil N, depleting soils for C3 competitors and producing the lowest soil N of any treatment at the end of the growth period (Wedin & Tilman 1993).

IIiii.) The evolutionary origins of C4 photosynthesis

The physiological and ecological characteristics of the C4 pathway described above have produced multiple hypotheses about what selective agent or agents may have favored the evolution of C4 species in different environments. Most of these have centered on carbon economy as the primary driver of C4 success, but there has also been speculation about the importance of aridity and seasonality, as well as the influence of ecological disturbances such as fire (Keeley & Rundel 2003, 2005). Reconstruction of paleoclimate in combination with fossil and molecular clock dating of C4 origins indicate that CO2 starvation was likely an important, but not exclusive factor in the proliferation of the C4 pathway (Ehleringer et al.

1991, 1997).

The oldest unequivocal C4 fossil is a petrified grass from the Ricardo Formation of

California, dated at 12.5 Ma (Tidwell & Nambudiri 1989). Molecular clock analyses

estimate that C4 photosynthesis first originated in Poaceae between 25-32 Ma (Christin et al.

2008a; Vincentini et al. 2008); C4 sedges (Cyperaceae) likely appeared later at 10.1-19.6 Ma

(Besnard et al. 2009). C4 eudicots are thought to have appeared more recently than C4 grasses, and molecular clock analysis has placed the first C4 eudicot in the family

Chenopodiaceae, between 14-21 Ma (Kadereit et al. 2003). While the accuracy of molecular clock analysis has been disputed (Pulquério & Nichols 2007), there remain no unequivocal fossils of C4 plants older than the mid-Miocene to verify these analyses, and the potential for fossil formation is particularly low in grasslands (Cerling 1999). Data from grass phytolith assemblages have been used to supplement the general lack of intact plant fossils. For example, phytolith data indicate the presence of Chloridoideae grasses, a mostly C4 clade, on the Great Plains of North America around 19 Ma (Stromberg 2005).

While C4 plants can trace their earliest origins to the Oligocene, C4 monocots did not come to dominate tropical and sub-tropical grasslands until the late-Miocene (5-10 Ma).

This conclusion is derived from analysis of the stable isotope composition of fossilized soils and animal remains. Because Rubisco preferentially fixes the light 12C isotope over the heavier 13C, plant tissues are enriched in 12C relative to the bulk atmosphere (Farquhar et al.

1989); however, C4 plants exhibit much lower rates of isotopic discrimination than C3 plants, and they consequently have higher amounts of 13C in their tissues and a higher value of δ13C

(O’Leary 1981). As such, when C4 grassland expansion took place in the late-Miocene, the

δ13C of soils and grassland herbivores increased dramatically. This phenomenon has been recorded separately in Pakistan, East Africa, central North America and Argentina and was notably absent from areas where C4 grasses are generally excluded today such as Western

Europe (Cerling et al. 1993; Quade & Cerling 1995; Cerling et al. 1999; Passey et al. 2002).

Thus, the considerable passage of time between the initial appearance of C4 photosynthesis in the Oligocene and the rapid, global expansion of C4 vegetation at low latitudes in the late-Miocene poses something of a quandary. The prevailing explanation for

C4 success was formulated by Ehleringer et al. (1991), who hypothesized that CO2 starvation in C3 plants was brought about by declining CO2 levels in the late-Miocene, favoring the more efficient C4 pathway of carbon fixation. The threshold of CO2 at which C4

-1 photosynthesis is favored over C3 is modelled to be near 550 µmol mol at 30°C in

-1 monocots and about 400 µmol mol in eudicots, and it falls to lower CO2 values at higher temperatures (Ehleringer et al. 1997). Thus, it was expected that when atmospheric CO2 concentration crossed this threshold, C4 plants would be favored. This hypothesis is useful in explaining the origin of C4 species in the Oligocene, as atmospheric CO2 values are indeed estimated to have fallen below 400-500 µmol mol-1 around 25-30 Ma, based on three independent proxies (Pagani et al. 1999; Pearson & Palmer 2000; Royer et al. 2001).

However, evidence is somewhat contradictory on whether a change in atmospheric CO2 occurred during the late-Miocene that might account for the prodigious expansion of C4 grasslands at that time, as estimates range from 200 to nearly 400 µmol mol-1 (Pagani et al.

1999; Royer 2006; Tripati et al. 2009).

To explain this phenomenon, then, several other facets of C4 physiological ecology have been invoked to determine why C4 grassland expansion took place in the late-Miocene and whether or not C4 expansion on different continents can be explained by similar factors.

In Pakistan, India and East Africa, an increase in the seasonality of rainfall is believed to have played a role as dry and wet seasons intensified (Keeley & Rundel 2003, 2005; Hopley et al. 2007; Osborne 2008). A substantial decrease in rainfall for a large part of the year may

22 have eroded the dominance of the arborescent life form in favor of graminoids, and the increased PWUE of C4 grasses may have also been a factor in this case. Increased seasonality may also have induced greater frequency of fires, as evidenced by a substantial increase in black carbon deposits in ocean sediments in the western Pacific Ocean during the late-Miocene (Keeley & Rundel 2003, 2005). An increase in fire frequency would generally promote the graminoid life form as well, and C4 grasses may recover more rapidly from burning than C3 species in open, high light environments (Knapp & Medina 1999). In North

America, explanations are less clear. The late-Miocene climate of central North America appears to have been relatively humid, and there is no evidence of increased seasonality or fire frequency during this time (Osborne & Beerling 2006). Temperature may have been a factor, as C4 grasses on the Great Plains achieved dominance south of 37°N around 7 Ma, but did not expand north of that latitude until the early Pliocene, about 4 Ma (Cerling et al. 1997;

Fox & Koch 2003, 2004). However, this period was generally characterized by somewhat lower, not higher, temperatures than the mid-Miocene (Zachos et al. 2001), and it is still unclear what triggered C4 grassland expansion in North America. Thus, it appears that low

CO2 concentration was likely a necessary precursor for C4 origins and eventual ecological dominance at low latitudes, but the late-Miocene expansion of C4 grasslands was probably influenced by a combination of other factors.

The evolution of C4 photosynthesis in the eudicots is generally regarded as a more recent phenomenon than in the monocots. This conclusion is partially derived from molecular clock analysis, which places the origin of C4 photosynthesis in Chenopodiaceae around 14-21 Ma and within Atriplex (Amaranthaceae) around 8-11.5 Ma (Kadereit et al.

2003). Empirical data are generally lacking from other lineages; however, the relative

paucity of C4 species among many eudicot genera relative to monocots (nine in Sesuvium; six in Anticharis and Zaleya; five in Gisekia and Flaveria; 2-3 in Mollugo; 2 in Allionia and

Okenia; 1 in Cypselea and Zygophyllum) implies a more recent origin than in the monocots

(Sage 2004). C4 eudicots in many lineages may not have appeared until the Pleistocene

-1 when atmospheric CO2 levels ranged from 180-280 µmol mol (Ehleringer et al. 1997).

CO2 concentrations that low would have provided an advantage for C4 eudicots over C3 at temperatures as low as 20°C (Ehleringer et al. 1997). This suggests improved carbon economy may have been the primary benefit of C4 photosynthesis in the eudicots, and assessments of photosynthetically intermediate species (described in the next section), which occur in several C4 lineages, may provide insight into the potential selective agents that advanced the evolution of the C4 pathway in these lineages.

III.) C3-C4 intermediate photosynthesis

IIIi.) The anatomy and physiology of C3-C4 photosynthetic plants

C3-C4 intermediate species were first described in 1974 by Kennedy & Laetsch in the eudicot Mollugo verticillata (Molluginaceae). They observed this species’ CO2 compensation point to be intermediate between C3 and C4 levels, indicating that it was reducing photorespiratory CO2 loss in some fashion. The specific mechanism responsible for this uniqueness was first characterized in 1988 by Rawsthorne et al. in Moricandia arvensis

(Brassicaceae) and by Hylton et al. (1988) in species of Moricandia, Mollugo, Flaveria and

Panicum (Poaceae). Carbon metabolism in these species is distinct from both the C3 and C4 pathways.

Among the 33 C3-C4 species described to date, all employ a “glycine shuttle” to reduce photorespiratory CO2 loss. In the photorespiratory pathway, CO2 is released when two molecules of glycine react to yield one CO2 and one serine, a reaction catalyzed by glycine decarboxylase (GDC) in the mitochondria. In C3 leaves, this reaction is ubiquitous in the mitochondria of chlorenchymatous leaf tissues, but in C3-C4 intermediate species, GDC in the mesophyll cells is lacking a crucial peptide subunit that is only expressed in bundle- sheath cells (Morgan et al. 1993). As such, glycine produced in the photorespiratory cycle accumulates in the mesophyll of C3-C4 leaves and diffuses down a concentration gradient into the bundle-sheath. Glycine decarboxylation can then be catalyzed in bundle-sheath mitochondria which, in C3-C4 leaves, are usually arranged centripetally along the interior of the bundle-sheath cell wall. This localizes photorespiratory CO2 loss to the interior of bundle-sheath cells. The bundle-sheath mitochondria in C3-C4 intermediates are typically overlain by a layer of chloroplasts such that, when CO2 is released in glycine decarboxylation, it can be immediately reabsorbed by chloroplasts and reassimilated by

Rubisco. While about 50% of photorespired CO2 can be recaptured in a C3 leaf, in C3-C4 intermediates, reassimilation exceeds 75% (Hunt et al. 1987; Monson & Rawsthorne 2000).

To accommodate the increased metabolic activity of the bundle-sheath, these cells are enlarged in C3-C4 intermediates relative to C3 levels and contain significantly higher numbers of mitochondria and chloroplasts, though generally not approaching the levels found in C4 bundle-sheath cells (Brown & Hattersley 1989).

The result of this spatial compartmentation of photorespiration is a reduction in photorespiratory CO2 loss from C3 levels. In most C3-C4 species, the glycine shuttle is the only mechanism by which photorespiratory CO2 loss is reduced, and these plants are termed

“Type I” intermediates (Edwards & Ku 1987). These species generally exhibit CO2 compenation points between 20-30 µmol mol-1 at 30°C (Monson & Rawsthorne 2000). In addition, some intermediates also exhibit somewhat elevated levels of PEPC activity and

CO2 fixation by PEPC, implying the function of a partial C4 cycle in these species. These are termed “Type II” intermediates (Edwards & Ku 1987), and their CO2 compensation points are roughly 10 µmol mol-1 at 30°C. The only known Type II species occur in the genus

Flaveria (F. floridana, F. ramosissima and F. anomala). Finally, several species of Flaveria

(F. brownii, F. palmeri, F. vaginata, and possibly F. haumanii) possess extremely high levels of PEPC activity, such that they exhibit CO2 compensation points near the C4 range of

-1 0-5 µmol mol at 30°C. These are termed “C4-like” intermediates, and they are mainly distinguishable from C4 species in that some Rubisco activity remains in the mesophyll tissue

(Reed & Chollet 1985; Moore et al. 1989).

The C3-C4 pathway results in a distinctive set of gas exchange characteristics. As noted, their CO2 compensation points are reduced from C3 levels. Also, they exhibit reduced sensitivity to O2 inhibition of photosynthesis. In C3 leaves, Γ increases linearly with O2 concentration, and Γ in C4 leaves is generally unaffected by increasing O2. C3-C4 species exhibit a characteristic biphasic or curvilinear response. An increase in O2 up to roughly 21-

35% will cause Γ to rise somewhat, but at lower rates than in C3. However, when O2 concentration exceeds this level, the bundle-sheath Rubisco becomes saturated with photorespired CO2, and further increases in O2 result in a significantly greater rate of increase in Γ, with a slope similar to that of the C3 response (Dai et al. 1996; von Caemmerer

2000). The slope of the Γ v. O2 response is lower in Type II intermediates than in Type I species (Holaday et al. 1984).

A few studies have examined the water-use and nitrogen-use efficiencies of C3-C4 species, with conflicting results. In the case of PWUE, surveys have generally used instantaneous measures of A/E to determine PWUE. Brown & Simmons (1980) reported higher PWUE in the C3-C4 species Panicum milioides at ambient and low CO2 than in the C3

Festuca arundinacea (Poaceae), while Apel (1994) determined that PWUE was unchanged from C3 levels in the C3-C4 species Flaveria pubescens, F. anomala and Moricandia arvensis. Kocacinar et al. (2008) conducted a comprehensive survey of A/E in Flaveria intermediates, and found no change from C3 levels in Type I and II species, while C4-like intermediates showed PWUE levels equivalent to C4. Monson (1989a) used a more robust measure of PWUE, the response of stomatal conductance to photosynthetic capacity, and found no significant difference between the C3 Flaveria cronquistii and the C3-C4 species F. floridana, F. pubescens and F. ramosissima. In the case of PNUE, conflicting results have also been reported. Bolton & Brown (1980) observed no difference in PNUE between the

C3-C4 Panicum milioides and the C3 Festuca arundinacea, but Monson (1989a) found 17% greater PNUE in the Type II species Flaveria ramosissima than in the C3 F. cronquistii, while PNUE was unchanged from C3 levels in the Type I species F. pubescens and the Type

II F. floridana. Together, these reports of PWUE and PNUE demonstrate considerable variation in these traits among C3-C4 species and leave unresolved the issue of how PWUE and PNUE are influenced by the degree of C4-cycle activity present in these species.

IIIii.) The taxonomic and ecological distributions of C3-C4 photosynthesis

There are 33 known C3-C4 species occuring in 11 genera, two in the Poaceae

(Neurachne with one species and Panicum with three) and nine in the eudicots. Among the

27 eudicots, there are up to 14 intermediates in Flaveria, 4 in Heliotropium, 3 in Moricandia, 2 each in Mollugo and Alternanthera and one in Parthenium (Brassicaceae), Chamaesyce

(Euphorbiaceae), Salsola (Amaranthaceae) and Cleome (Cleomaceae) (Edwards & Ku 1987;

Voznesenskaya et al. 2001; Vogan et al. 2007; Voznesenskaya et al. 2007; RF Sage & MW

Frohlich, unpublished). In addition, there has not been an exhaustive analysis of all C4 lineages for the presence of C3-C4 intermediate species, and there may be others that have not yet been characterized. For example, three species of Heliotropium (H. lagoense, H. filiforme and H. cremnogenum) have Kranz-like leaf anatomy (Frohlich 1978), but they are

13 clearly not C4 species given their C3-like δ C values (RF Sage & MW Frohlich, unpublished), and they may also be C3-C4 intermediate species.

The extraordinary abundance of intermediate species in the relatively small Flaveria genus (23 species) and the diversity of physiological traits among them have made it the most thoroughly-studied system in C3-C4 biology. The intermediate physiology of Flaveria species led to hypotheses that this pathway represented an intermediate state in the evolution of the C4 pathway (Monson 1989b; Monson & Moore 1989; Monson 1996), and a recently published, species-level phylogeny shows that this is indeed the case (McKown et al. 2005).

While other taxa lack such comprehensive analysis, it is reasonable to expect that this may also have been the case for many other C4 lineages containing C3-C4 species. One factor supporting this hypothesis is that the characteristic partitioning of GDC into the bundle- sheath cells found in all known C3-C4 species is also present in all C4 species analyzed to date. These include species of all three C4 biochemical subtypes (NADP-ME, NAD-ME and

PCK) from both monocot genera (Sorghum, Zea, Chloris, Panicum, Eriachne, Setaria) and eudicots (Flaveria and Amaranthus) (Ohnishi & Kanai 1983; Farineau et al. 1984; Hylton et

al. 1988; Morgan et al. 1993; Ueno 2001; Yoshimura et al. 2004). Though C4 species experience very low rates of RuBP oxygenation, the photorespiratory pathway in these species must proceed through glycine decarboxylation in bundle-sheath mitochondria in an identical fashion to C3-C4 intermediates.

While C3-C4 species have been found in multiple separate evolutionary lineages, they generally share common features in their ecological distribution in open, high light environments, generally at low latitudes. For example, Salsola arbusculiformis is found in open, calcareous soils on low mountain slopes of the Kyzylkum Desert in Central Asia

(Kapustina 2001). In East Africa, Cleome paradoxa is found in open, seasonally dry habitats

(Voznesenskaya et al. 2007). In the Western Hemisphere, Mollugo verticillata is common along roadsides and disturbed areas in North America (Hickman 1993). C3-C4 species of

Heliotropium and Chamaesyce are present in sparsely vegetated, arid, sandy soils of the southwestern United States and northern Mexico (Lundell 1966; Correll & Johnston 1970;

Webster et al. 1975; Frohlich 1978). Flaveria intermediates are generally found on alkaline or gypseous soils in open, high light environments in Mexico, Florida, the Caribbean and the southwestern United States (Powell 1978; McKown et al. 2005). This affinity for open, high light environments in low latitudes likely indicates that C3-C4 species are generally favored by the same factors that influence C4 success - namely high rates of photorespiration brought about by high leaf temperatures and, in several cases, aridity.

IV.) Models of C4 evolution

The multiple origins of C4 and C3-C4 photosynthesis in taxa such as Flaveria,

Cyperaceae and Poaceae subfamilies Aristoideae, Panicoideae and Chloridoideae implies

29 that there may be some common set of anatomical, physiological and/or genetic characters that predisposes a lineage towards evolving components of C4 metabolism (Monson &

Rawsthorne 2000; Sage 2004; Christin & Besnard 2009). For example, enlarged bundle- sheath cells may be required to facilitate subsequent enhancement of bundle-sheath organelle content and metabolic activity (Monson & Rawsthorne 2000). To determine whether or not there have been common preconditions to C4 evolution in multiple lineages, it is necessary to combine anatomical and physiological assessments of these species with phylogenetic data to ascertain whether C3 and C3-C4 species that are closely-related to C4 taxa possess characteristics that may have facilitated the evolution of C4 traits. Because the evolution of this complex metabolic pathway is thought to have occurred in several steps (Edwards & Ku

1987; Brown & Hattersley 1989; Rawsthorne 1992; Monson 1996), each stage in C4 evolution would need to enhance plant fitness in its own right, or at least be neutral. Given the intermediate phylogenetic position of certain C3-C4 species in Flaveria and Alternanthera

(McKown et al. 2005; Sanchez-del Pino 2009), an understanding of the particular advantages that the C3-C4 pathway confers in the warm, often arid environments where these species occur may prove especially important in evaluating the sequence of events in C4 evolution and the relative contribution of each stage to plant performance and fitness.

In terms of leaf anatomy, C4 plants are generally characterized by higher vein density and lower mesophyll-to-bundle sheath ratios than C3 species (Dengler & Nelson 1999). This is likely a necessary alteration for C4 species since the transport of C4 metabolites beween mesophyll and bundle-sheath cells would be greatly enhanced by shorter diffusional distances (Dengler & Nelson 1999). It has been speculated that increased vein density may be a necessary precondition for the evolution of C4 metabolism and may be present in C3

species prior to the evolution of C4 traits (Sage 2004). Specifically, increased vein density may ameliorate evaporative stress for C3 species in arid environments by reducing the ratio of evaporative surface area to that of water-conducting tissue (Roth-Nebelsick et al. 2001) and potentially improving water storage capacity as a buffer against wind gusts (Sage 2004).

In Flaveria, there is evidence to suggest that vein density is increased in the C3 F. robusta relative to the other C3 species in the genus (McKown & Dengler 2007). This result is of note because F. robusta is sister to the C4- and C3-C4-containing clade of Flaveria (McKown et al. 2005), and the presence of higher vein density in F. robusta may indicate that this trait directly preceded the evolution of C3-C4 intermediate photosynthesis in the genus (McKown

& Dengler 2007, 2009). Enlarged bundle-sheath cells may also be an important precondition to C4 evolution, and in C3 species this trait may enhance water storage capacity (Sage 2004).

In Flaveria, the bundle-sheath cells of C3 species are somewhat smaller and less elongated than C3-C4 and C4 species which may indicate that an increase in bundle-sheath cell size does not occur until the introduction of larger numbers of organelles to the bundle-sheath in C3-C4 intermediates (McKown & Dengler 2007). However, in Heliotropium, C3 species such as H. tenellum and H. karwinskyi, which are likely sister to the C4 and C3-C4-containing-clade of the genus (Frohlich et al., unpublished), do have larger bundle-sheath cell volumes than H. europeaeum, which is not closely related to the C4 taxa (Muhaidat 2007). Larger bundle- sheath cells that comprise a higher proportion of leaf volume may necessarily assume larger numbers of organelles, especially chloroplasts, to avoid reducing overall leaf photosynthetic capacity as bundle-sheath cell size increases (Sage 2004), and an increase in C3 bundle- sheath cell size and organelle allocation may be another important precondition in ancestral

C3 species to evolving C4 traits, though analysis from more taxa is needed to determine if this is a recurrent pattern in multiple evolutionary lineages.

Another early precondition to C4 evolution may be the duplication of important C4 genes. Gene duplication would permit novel mutations to appear in duplicate genes while still maintaining necessary physiological functions in redundant gene copies, and this process may be necessary in providing the raw material for new variations in metabolically important genes (Monson 1999, 2003). There is evidence to suggest that such gene duplication has occurred in Poaceae. In the Aristidoideae subfamily, there have been two independent origins of C4 photosynthesis, and distinct PEPC gene lineages appeared through several rounds of gene duplication before or early during grass diversification (Christin & Besnard

2009). In the Chloridoideae and Panicoideae subfamilies, there have also been multiple origins of C4 photosynthesis, and there may have been up to five separate duplications of the gene coding the decarboxylating enzyme PCK in these groups (Christin et al. 2008c).

Annual or short-lived perennial species may be relatively more disposed towards gene duplication since they reproduce on relatively short times scales and purifying selection may remove deleterious genetic alterations rapidly while allowing advantageous ones to spread

(Monson 2003).

If the basic anatomical and genetic preconditions described above are present in ancestral C3 species, the subsequent steps towards evolving the C4 pathway would primarily entail changes in the localization and expression of metabolically important proteins, traits that are prominent characteristics of C3-C4 intermediate species. In Type I C3-C4 intermediates, a glycine shuttle system is functional if a single, crucial peptide subunit of glycine decarboxylase is not functional in the mesophyll, such as in Moricandia arvensis

(Brassicaceae) (Morgan et al. 1993). Photorespiratory metabolites accumulate in the mesophyll and diffuse into the bundle-sheath where a catalytically competent GDC is present to continue the photorespiratory cycle and localize photorespiratory CO2 emission to the bundle-sheath. Increased vein density would facilitate such metabolite diffusion. This single genetic change in GDC expression may effect the primary alterations to bring a C3 species with the necessary preconditions to being a Type I C3-C4 intermediate. Subsequently, enhanced expression of PEPC in the mesophyll, as is found in Type II intermediates like

Flaveria floridana, F. ramosissima and F. anomala, may be beneficial at that point to recapture CO2 that diffuses out of the bundle-sheath, and a partial C4 cycle would begin to form in this case (Monson et al. 1986). In tobacco, enzymes such as NADP-ME, NAD-ME,

PCK and PPDK exhibit activities 9 to 18 times greater in bundle-sheath cells than mesophyll cells (Hibberd & Quick 2002), and high expression levels of these enzymes in the bundle- sheath could further facilitate the formation of a partial C4 cycle. As PEPC assumes a more prominent role in carbon assimilation and begins to compete directly with Rubisco for CO2 molecules, strong enzyme partitioning, especially localization of Rubisco to the bundle- sheath, could improve integration between the C3 and C4 cycles and further improve carboxylation efficiency and plant carbon balance (Monson & Rawsthorne 2000; Sage 2004).

C4-like intermediates display this strong enzymatic segregation, with over 90% of leaf

Rubisco activity localized to bundle-sheath cells in C4-like intermediates such as Flaveria brownii and F. palmeri (Monson et al. 1987; Moore et al. 1989) compared to less than 30% in the Type II intermediate F. ramosissima (Moore et al. 1988). Once most of the leaf’s

Rubisco is isolated to bundle-sheath cells and the C4 pathway accounts for the substantial majority of initial carbon assimilation, modifications that optimize C4 cycle function may

then be selected for. For example, C4 Rubisco generally has lower CO2-specificity than C3

Rubisco because it operates in a high CO2 environment (Kubien et al. 2008), and this change has been recorded in several independent lineages (Christin et al. 2008b). Such alterations in

Rubisco specificity were not observed in C3-C4 Flaveria species (Kubien et al. 2008). Also,

PEPC and NADP-ME from C4 leaves are less inhibited by malate than those from C3 leaves

(Drincovich et al. 2001; Svensson et al. 2003), and this would also improve the plant’s ability to operate an effective C4 cycle producing large amounts of malate without experiencing substantial enzymatic inhibition. Increased bundle-sheath cell wall thickness would also be important at this stage for maintaining high CO2 concentrations around

Rubisco, and, in Flaveria, thicker bundle-sheath walls are apparently present primarily in the

C4-like and C4 species (McKown & Dengler 2007).

The importance of increased PWUE and PNUE to the success of C3-C4 intermediates is somewhat unclear. While these traits would likely enhance plant fitness (Monson 1989a), the presence of a partial C4 cycle in Type II C3-C4 Flaveria species is not definitively linked to higher PWUE and PNUE (Monson 1989b, Kocacinar et al. 2008), and further research is necessary to ascertain the extent of these traits in C3-C4 species, their correlation to C4-cycle activity, and their relative importance in C3-C4 plant performance.

V.) Thesis objectives

The model of stepwise C4 evolution described above is based primarily on data from the Flaveria system; however, it is necessary to develop alternative systems to Flaveria to assess the particular path, or paths, that led to C4 photosynthesis in different plant lineages and the particular advantages of each putative intermediate stage to plant performance and

fitness. Also, a detailed analysis of carbon balance and resource-use efficiency in C3-C4 intermediates and closely related C3 and C4 species is required to determine the contribution of C3-C4 traits to plant performance, particularly in the low CO2, high temperature conditions that are thought to have been most important in selecting for C3-C4 traits and advancing the initial stages of C4 evolution (Ehleringer et al. 1991, 1997; Sage 2004).

The second chapter of this thesis addresses for the first time the physiological intermediacy of species in the genus Heliotropium (Boraginaceae). This includes an evaluation of photosynthetic gas exchange traits commonly found in C3-C4 species such as reduced CO2 compensation point and reduced sensitivity to O2 and a diagnosis of the extent of photosynthetic intermediacy present in these species. It also includes an evaluation of leaf-level photosynthetic performance under low CO2 conditions and the implications for these species’ performance in their natural habitats. This chapter has been published as

Vogan, Frohlich & Sage (2007) in Plant, Cell and Environment.

The third chapter of this thesis address the water-use and nitrogen-use efficiencies of

C3-C4 species of Flaveria. This includes a robust measure of PWUE and PNUE that has not previously been applied comprehensively across the majority of species in the genus. This study provides information about the influence of C4-cycle activity on these ecophysiological traits. It also includes an evaluation of the evolution of stomatal mechanism in C3-C4

Flaveria species and an assessment of plant nitrogen relations, both of which may be important in natural habitats.

The fourth chapter of this thesis addresses the performance of different

-1 photosynthetic types under low CO2 (180 µmol mol ) and high temperature (37/29°C day/night) conditions. This study includes a C3, C4 and C3-C4 species from each of three

eudicot genera (Heliotropium, Flaveria and Alternanthera) and is the first analysis of C3-C4 photosynthetic performance under the conditions that are postulated to have selected for their evolution. This includes an evaluation of photosynthetic and stomatal acclimation to low

CO2 concentrations in the different photosynthetic types as well as the responses of carbon assimilation, PWUE and PNUE to high temperature and low CO2.

Chapter Two - The functional significance of C3-C4 intermediate traits in Heliotropium

L. (Boraginaceae): gas exchange perspectives

Abstract

In this study, I demonstrated for the first time the presence of species exhibiting C3-

C4 intermediacy in Heliotropium (sensu lato), a genus with over 100 C3 and 150 C4 species.

CO2 compensation points (Γ) were intermediate between C3 and C4 values in three species of

-1 Heliotropium: H. convolvulaceum (Г= 20 µmol CO2 mol air), H. racemosum (Г= 22 µmol mol-1 ), and H. greggii (Г= 17 µmol mol-1). Photosynthetic water-use efficiencies (PWUE) were also somewhat elevated from C3 levels in H. convolvulaceum and H. racemosum.

Heliotropium procumbens may also be a weak C3-C4 intermediate based on a slight reduction

-1 -1 in Γ (48.5 µmol CO2 mol ) compared to C3 Heliotropium species (52 to 60 µmol mol ).

The intermediate species H. convolvulaceum, H. greggii and H. racemosum exhibited over

-1 50% enhancement of net CO2 assimilation rates at low CO2 levels (200-300 µmol mol ); however, no significant differences in stomatal conductance were observed between the C3 and C3-C4 species. I also assessed the response of Г to variation in O2 concentration for these species. Heliotropium convolvulaceum, H. greggii and H. racemosum exhibited similar responses of Γ to O2 with response slopes that were intermediate between the

-1 responses of C3 and C4 species below 210 mmol O2 mol air. The presence of multiple species displaying C3-C4 intermediate traits indicates that Heliotropium could be a valuable new model for studying the evolutionary transition from C3 to C4 photosynthesis.

Introduction

36 37

C4 photosynthesis is a CO2-concentrating mechanism that suppresses photorespiration and enhances the carboxylation efficiency of Rubisco. As a consequence, photosynthetic capacity, water-use efficiency, nitrogen-use efficiency, and biomass productivity are enhanced in warm environments where photorespiration is substantial in C3 plants.

Humanity’s most productive crops are C4 plants, and the leading new sources of biofuels use the C4 pathway (Brown 1999; Samson et al. 2005). Because of its high productive potential, the introduction of the C4 pathway into C3 crops such as rice is seen as an important means of meeting the nutritional demands of an enlarged human population later this century (Dawe

2000). To engineer C4 crops, however, major hurdles must be overcome and there is great uncertainty in how to proceed, particularly with regards to the structural changes required to compartmentalize C3 and C4 metabolism (Leegood 2002). For insights on how to engineer

C4 photosynthesis into C3 species, we can exploit the many natural lineages that independently evolved the C4 pathway. C4 photosynthesis has evolved over 40 times in 19 families of monocots and dicots (Muhaidat et al. 2007). In a number of these evolutionary lines, vestiges of the evolutionary transition potentially remain in species that exhibit intermediate traits between fully developed C3 and C4 photosynthesis (Sage 2004; McKown et al. 2005).

The group that has contributed the most to our understanding of C4 evolution is the genus Flaveria (Asteraceae) (Monson 1996). Flaveria contains 23 species, of which 12 express intermediate traits that range from C3-like to C4-like forms (McKown et al. 2005).

Some of the C3-C4 intermediate species concentrate CO2 into the bundle sheath exclusively by the restriction of glycine decarboxylation to the bundle sheath compartment (Monson &

Rawsthorne 2000). This is termed Type I C3-C4 intermediacy (Edwards & Ku 1987) but can

also be regarded as “C2 photosynthesis” because the decarboxylation of 2-carbon photorespiratory metabolites concentrates CO2 around Rubisco, enhancing photosynthetic efficiency. A few species also express a pattern of intermediacy where a limited C4 cycle is present in addition to the localization of glycine decarboxylase, as indicated by elevated activity of PEP carboxylase and C4 decarboxylating enzymes; this is termed Type II intermediacy (Edwards & Ku 1987). With the recent publication of a detailed, species-level phylogeny of Flaveria based on molecular characters (McKown et al. 2005), it is now possible to interpret C3-C4 intermediacy in an evolutionary context. In the case of Flaveria, only two species expressing intermediate characteristics occur in the evolutionary lineage leading from C3 to C4 photosynthesis; most of the intermediates occur in a separate clade

(McKown et al. 2005). One of these two intermediate species, F. ramosissima, expresses

Type II intermediacy as indicated by partial C4 cycle engagement; the other, F. palmeri, is

C4-like with a nearly complete C4 cycle (Ku et al. 1991). With this limited number of evolutionary intermediates in Flaveria, the ability to draw evolutionary inferences is weakened, and it is thus necessary to develop a broader approach that can exploit comparative methods from evolutionary biology (Harvey & Pagel 1991). For a comparative approach, multiple model systems are needed. Unfortunately, other genera in which C3-C4 intermediacy occurs have limited utility in the study of C4 evolution. Some genera have only one or two confirmed intermediate species (Brassica, Alternanthera, Parthenium,

Neurachne, Salsola) (Apel, Horstmann & Pfeffer 1997; Sage, Li & Monson 1999;

Voznesenskaya et al. 2001), all lack a well-resolved phylogeny, and some clades with C3-C4 intermediates lack C4 species altogether (Moricandia, Diplotaxis) (Apel et al. 1997).

One promising genus is Heliotropium (Boraginaceae), whose systematics were characterized by Frohlich (1978) and more recently by Hilger & Diane (2003). C4 photosynthesis in the genus Heliotropium occurs in section Orthostachys (= Euploca sensu

Hilger & Diane 2003), a group with approximately 140 species including dozens of C3 species and a number of potential C3-C4 intermediates (Frohlich 1978). In addition, section

Orthostachys contains numerous C3 species with “Kranz-like” anatomy, characterized by enlarged bundle sheath tissues and reduced mesophyll volume (Frohlich 1978). The large number of species in section Orthostachys indicates Heliotropium could be an excellent model to complement Flaveria in comparative analyses of C2 photosynthesis and its relevance to C4 evolution. Characterization of potential evolutionary intermediates and their

C3 relatives in Heliotropium would facilitate comparative analyses that address such issues as whether C4 evolution is constrained to follow similar steps and whether there are common initial preconditions that favor the evolution of the C4 pathway.

In recent years, this lab group has collected over a dozen living Heliotropium species from their native environments. Here I characterize the gas exchange response of ten species that are putatively placed on or near the transition from the C3 to the C4 pathway in recent phylogenetic assessments (Frohlich, 1978 and unpublished; Hilger & Diane 2003). I assessed carbon isotope ratios, water-use efficiencies, the CO2 response of photosynthesis and stomatal conductance, and the response of the CO2 compensation point to variation in

O2. With these data, I evaluate the functional significance of C2 photosynthesis in

Heliotropium.

Materials and Methods

Seed sources and growth conditions

Seeds and cuttings were obtained between 2001 and 2005 from sites in Mexico,

Jordan and the United States as summarized in Table 2.1. All plants except H. europaeum are placed in Heliotropium section Orthostachys; H. europaeum occurs in section Heliotropium which is entirely C3 (Frohlich 1978; Hilger & Diane 2003; Sage & Frohlich, unpublished results from a carbon isotope survey). I included H. europaeum in this study in order to ensure a pure C3 species was present in case all putative C3 species of section Orthostachys exhibited incipient C3-C4 intermediacy.

Plants used for response curves of net CO2 assimilation rates (A) vs. intercellular CO2 concentrations (Ci) and assessments of photosynthetic water-use efficiency (PWUE) were grown from seed in the University of Toronto greenhouse during May to September, 2005.

The approximate day/night temperature of the greenhouse was 30°C/25°C and peak midday irradiance exceeded 1600 µmol photons m-2 s-1 on sunny days. Plants were watered as needed to avoid drought stress, and were fertilized twice weekly with 1.4 g L-1 each of 15-

30-15 Miracle Gro (Scotts Canada, Mississuaga, Ontario, Canada) and 30-10-10 Plant-

Products Evergreen Fertilizer (Brampton, Ontario, Canada) supplemented with micronutrients. I found that this regime produced healthy plants with no symptoms of nutrient deficiency, while more conventional nutrient regimes such as Hoagland’s solution or any single commercial brand produced plants with deficiency symptoms.

Gas exchange and carbon isotope ratios

Photosynthetic gas exchange was measured using an LI-6400 portable photosynthesis system (Li-Cor, Inc., Lincoln, NE, USA) during June to September and between 0900 and

1700 hours. Measurements were conducted at a leaf temperature of 30°C, a photosynthetic photon flux density of 1200 µmol m-2 s-1, and leaf-to-air vapor pressure deficit (VPD) of 1.8 kPa. Gas exchange was measured two to four months after planting using the most recent fully-expanded leaf. All measurements used the standard cuvette (LI 6400-02) except for measurements of H. greggii which used the conifer chamber (LI 6400-05). To minimize gas exchange in the photosynthetically active stem, the stems of H. greggii were coated with silicone-based high vacuum grease (Dow Corning). CO2 compensation points were calculated as the x-intercept of a linear regression through the five lowest intercellular CO2 values on a graph of net CO2 assimilation vs. Ci. The initial point in each A/Ci curve was

-1 logged at 370 μmol CO2 mol air . CO2 was then lowered to 300, 200, 120, 80, 50, and 30

μmol mol-1, then raised to ambient again and allowed to stabilize at the original values of A

-1 and stomatal conductance (gs). The CO2 level was then raised to 600 and 800 μmol mol to assess the maximum rate of photosynthesis. In a separate set of measurements from the A/Ci analyses, water-use efficiency was determined at ambient CO2 in identical conditions, except that VPD was 2.0 kPa.

O2 sensitivity of the CO2 compensation point (Г) was measured on plants grown from seed in growth chambers (Model GC-20, Enconair Ecological Chambers, Winnipeg, MB,

R2W 3A8, Canada) at 30/22°C day/night temperatures with irradiance between 500 and 600

µmol photons m-2 s-1 during the photoperiod. Plants were fertilized as described for greenhouse-grown plants and measured using an LI-6400 photosynthesis system coupled with a gas mixer that delivered O2 concentrations of 20, 50, 100, 150, 210, 350, and 500 mmol mol-1. Gas concentration was established using two mass flow controllers (Model 840,

Sierra Instruments, Monterey, California, USA) that regulated the mixing ratio of pure N2

and O2. CO2 was supplied and mixed to desired concentrations via the LI-6400 console.

CO2 compensation points were determined as described above.

Carbon isotope ratios of Heliotropium leaves were measured by the University of

California at Davis stable mass isotope facility (http://stableisotopefacility.ucdavis.edu/).

Plants used for carbon isotope ratios were sampled in the greenhouse concurrent with A/Ci measurements. Plants were vouchered and stored in the herbarium of the Royal Ontario

Museum, Toronto, Ontario, Canada.

Statistical analysis

Species values for CO2 compensation points and water-use efficiencies were compared using a one-way ANOVA. Both data sets were normally distributed with equal variance (P > 0.05), allowing for pairwise comparisons using Tukey’s HSD test. Species

-1 values for A at 370 and 800 µmol mol (A370 and A800), stomatal conductance at 370 µmol

-1 mol (g370) and initial slope of the A/Ci response curves lacked normality and were instead compared using a one-way ANOVA on ranks with Dunn’s test used for pairwise comparisons. Due to small sample size, statistical power was limited, hindering the ability to resolve species-level differences in the non-parametric pairwise tests. To improve statistical power in resolving differences between photosynthetic functional types, the data for initial

A/Ci slopes, stomatal conductances, Ci/Ca ratios, and photosynthetic rates were pooled according to photosynthetic type and compared using a nested one-way ANOVA; pairwise comparisons were conducted using Tukey’s HSD test. The slopes of the Г vs. O2 response curves were compared between functional types for measurements between 20 and 210

-1 mmol mol O2, which corresponded to the first of two apparent linear phases in the C3-C4

43 response curves. Because statistical power was again low, I pooled the slope values between

20 and 210 mmol O2 by photosynthetic type and compared them using a nested one-way

ANOVA with Tukey’s HSD test used for pairwise comparisons. The apparent change in slope among the intermediate species at 210 mmol O2 was assessed for statistical significance by fitting linear models to the two apparent linear phases of the response curve and comparing the two slopes using an F-ratio test (Ott 1977).

Results

The carbon isotope ratios for all of the species studied except H. texanum were between -25.3‰ and -30.6‰ (Table 2.2). Heliotropium texanum had a typical C4 isotopic ratio of -14.7‰.

Heliotropium texanum exhibited a typical CO2 response of A for a C4 plant, as evidenced by a low CO2 compensation point, a steep initial slope of the A vs. Ci response, and a sharp CO2 saturation point (Fig. 2.1A). The A vs. Ci response in H. europaeum was typical for a C3 species, as demonstrated by the relatively high CO2 compensation point of 52

-1 μmol mol , a shallow initial slope, and a gradual rise of A to a CO2 saturation point above

500 µmol mol-1 (Fig. 2.1B). Similarly, H. tenellum, H. calcicola, H. fallax, H. karwinskyi and H. procumbens also exhibited A/Ci responses that were typical of C3 species (Fig. 2.1A,

2.1B). The A/Ci responses in H. greggii, H. racemosum, and H. convolvulaceum were C3- like, except that the x-intercept was shifted to lower values. The initial slope and A800 in H. greggii were equivalent to initial slope and A800 values observed in the C3 species H. karwinskyi and H. europaeum (Fig. 2.1B; Table 2.2). Heliotropium convolvulaceum and H.

racemosum appeared to have higher initial slope and A800 values than H. greggii and the C3 species (Fig. 2.1); however, these differences were not statistically significant.

The x-intercept of the A/Ci response was used to determine Г (Table 2.2). The C3 species H. europaeum exhibited a Γ of 52 μmol mol-1, which is typical of this pathway. The

Г values of H. tenellum, H. calcicola, H. fallax, and H. karwinskyi were also within the C3 range at 55, 60, 54, and 52 μmol mol-1, respectively. In H. procumbens, Г was slightly less

-1 than its C3 congeners at 48.5 μmol mol . The values for H. convolvulaceum and H. racemosum were 20 μmol mol-1 and 22 μmol mol-1, respectively, and H. greggii had a Γ of

-1 -1 17 μmol mol . In H. texanum, Г was 5.0 μmol mol , typical of C4 photosynthesis.

The C4 H. texanum had the highest water-use efficiency, followed by H. convolvulaceum, H. racemosum, and H. greggii, whose PWUE values were somewhat elevated from C3 levels, significantly so in the cases of H. convolvulaceum and H. racemosum (Table 2.2). Heliotropium procumbens showed no significant difference in

PWUE from the C3 species. Absolute stomatal conductance values were not significantly different between the species (Table 2.2); however, differences in gs could reflect variation in

A and reduce statistical resolution (Farquhar & Sharkey 1982). To minimize variation associated with differing A, I normalized gs to the corresponding maximum gs observed at low CO2 (Table 2.5). No significant difference in normalized gs was observed between species (not shown), nor between photosynthetic functional types when the normalized gs was pooled and compared at CO2 levels of the past and present (Table 2.3). Mean A at 200,

-1 300, and 370 µmol mol was significantly greater in the C3-C4 than C3 species (Table 2.3).

Because of the greater A and similar conductance values observed, Ci/Ca was reduced in the

C3-C4 species relative to the C3 species (Table 2.3).

In the assessment of the Γ vs. O2 response, Γ in H. texanum showed no response to

-1 varying O2 between 20 and 500 mmol mol (Fig. 2.2). The response in the C3 species H. europaeum, H. tenellum, and H. karwinskyi was typical of C3 responses in that Γ linearly

-1 increased with rising O2 from 20 to 500 mmol mol (Fig. 2.2; Table 2.4). By contrast, the response of Γ vs. O2 in H. convolvulaceum, H. racemosum, and H. greggii showed two

-1 apparent phases. Below an O2 level of 210 mmol mol , the mean slope of Γ vs. O2 in these

-1 three species was 1.86 mmol CO2 mol O2, a third less than the mean slope value observed

-1 in the C3 species (Table 2.4). Around 210 mmol mol , a significant change in slope (F(3,122) =

142.2, P = 0.016) was observed among the C3-C4 species, and the slope of the second linear

-1 phase (between 210 and 500 mmol mol ) increased in the C3-C4 species to values similar to that observed in the C3 species (Fig. 2.2).

Species Habitat Collection site H. europaeum L. Weedy species occurring in Muhaidat sn. Irbid, Jordan disturbed areas and pastures, semi- 32°33' N 35°51' E arid and arid regions. H. calcicola Fernald Mature community on undisturbed Frohlich 2522. Ridge crest hillsides, limestone deserts near Reforma, Mexico 23°36' N 99°17' W H. fallax I.M. Johnst. Rocky, brushy mountainside Frohlich 2501. San Juan La Jarcia, Mexico 16°27' N 95°51' W H. tenellum (Nutt.) Torr. Open rocky or gravelly soils, Vogan sn. Victoria Glades limestone glades Conservation Area, MO 38°12' N 90°33' W H. karwinskyi I.M. Johnst. Mature, undisturbed communities Frohlich 2521. Ridge crest on limestone, semi-arid zones near Reforma, Mexico 23°36' N 99°17' W H. procumbens Miller Weedy species, highly disturbed Frohlich 1538. Oaxaca, soils Mexico 17°07' N 96°72' W H. convolvulaceum Semi-stable sand dunes in arid Vogan sn. Boquillas Canyon, (Nutt.) Gray zones TX, USA 29°13' N 102°55' W H. racemosum (Rose & Sandy places on the coastal plain of Vogan sn.Mustang Island, TX, Standl.) I. M. Johnston southern Texas USA 27°45' N 97°07' W H. greggii Torr. Roadside shoulders, sand, gravel, Vogan sn. El Paso, TX, USA clay flats, arid and semi-arid soils 31°45' N 106°30' W H. texanum I.M. Johnst. Semi-arid sandy soils of the Rio Vogan sn. Zapata, TX, USA Grande Valley 26°57' N 99°17' W

Table 2.1 Habitat and collection information for Heliotropium species in this study. Habitat

information from Johnston (1924), Correll & Johnston, (1970), Frohlich (1978), and field notes

of the collecting individuals.

13 -1 Species δ C N= Γ (µmol PWUE Initial slope of A at 370 A at 800 gs at 370 μmol mol -1 -1 -1 (‰) mol ) (µmol A/Ci response μmol mol μmol mol external CO2 CO2 mmol (ΔA/ΔCi) external external -1 H2O ) CO2 CO2 H. europaeum -28.1 5 52.4 ± 1.2ab 3.0 ± 0.2a 0.151 ± 0.021a 21.6 ± 1.7ab 35.3 ± 2.1ab 0.380 ± 0.071a H. calcicola -25.3 4 60.0 ± 1.7a 3.9 ± 0.1ab 0.135 ± 0.013a 22.8 ± 2.1ab 29.9 ± 4.2ab 0.301 ± 0.021a H. fallax -26.9 5 53.8 ± 3.3a 3.0 ± 0.4a 0.109 ± 0.006a 19.5 ± 0.6a 28.7 ± 1.3ab 0.345 ± 0.042a H. tenellum -30.1 5 55.2 ± 1.1a 2.6 ± 0.1a 0.121 ± 0.005a 21.3 ± 0.8a 27.1 ± 1.7a 0.420 ± 0.050a H. karwinskyi -30.3 2 52.1 ± 0.0ab 3.4 ± 0.1ab 0.151 ± 0.038a 21.4 ± 2.7ab 35.7 ± 3.3ab 0.315 ± 0.077a H. procumbens -28.3 6 48.5 ± 1.0b 2.9 ± 0.2a 0.131 ± 0.011a 24.4 ± 1.0ab 30.3 ± 0.4ab 0.438 ± 0.036a H. convolvulaceum -30.6 5 19.8 ± 0.8c 5.0 ± 0.5c 0.176 ± 0.014a 30.1 ± 4.6b 41.5 ± 6.0ab 0.314 ± 0.114a H. racemosum -30.4 3 21.9 ± 0.6c 4.8 ± 0.5c 0.175 ± 0.046a 26.5 ± 3.7ab 42.3 ± 5.8ab 0.291 ± 0.034a H. greggii -28.7 5 17.3 ± 1.1c 4.2 ± 0.3bc 0.144 ± 0.009a 26.8 ± 1.5ab 33.6 ± 1.3ab 0.329 ± 0.028a H. texanum -14.7 5 5.0 ± 0.4d 6.6 ± 0.3d 2.01 ± 0.53b 42.7 ± 2.4c 43.6 ± 2.4b 0.315 ± 0.019a

Table 2.2

13 Table 2.2. δ C values, CO2 compensation points (Γ), photosynthetic water-use efficiencies

(PWUE), and data derived from A/Ci responses shown in Figure 2.2, including the initial slope

and stomatal conductance (gs) at current CO2 levels for the Heliotropium species in this study.

Data are mean ± SE. Superscripts indicate significant difference at P < 0.05 using the Tukey

13 test. δ C sample size was two for each species. CO2 compensation points were determined

from the x-intercept of A/Ci curves taken at 30°C, vapor pressure deficit of 1.8 kPa and

photosynthetic photon flux density of 1200 μmol photons m-2 s-1. Water-use efficiency was

measured at 370 µmol mol-2 and a vapor pressure difference between leaf and air of 2.0 kPa.

Samples size for gas exchange measurements were 2-6 as indicated.

Photosynthetic N Initial slope Net CO2 assimilation rate at varying Normalized stomatal conductance at Ci/Ca ratio at -1 -1 type of A/Ci external CO2 (µmol mol ) varying external CO2 (µmol mol ) 370 µmol -1 response mol CO2 (ΔA/ΔCi) 200 300 370 200 300 370 a a a a a a a a C3 21 0.12 ± 0.01 10.3 ± 0.5 16.8 ± 0.6 21.8 ± 0.6 0.46 ± 0.05 0.47 ± 0.05 0.45 ± 0.05 0.70 ± 0.01

a b b b a a a b C3-C4 13 0.15 ± 0.01 16.4 ± 1.1 23.1 ± 1.6 28.8 ± 1.9 0.48 ± 0.06 0.44 ± 0.06 0.43 ± 0.06 0.56 ± 0.02

b c c c a a a c C4 5 1.5 ± 0.1 32.2 ± 1.5 40.3 ± 2.2 41.1 ± 2.4 0.35 ± 0.02 0.34 ± 0.02 0.32 ± 0.02 0.40 ± 0.03

Table 2.3. Initial slopes of A/Ci responses, net CO2 assimilation rates, relative stomatal conductances, and ratios of intercellular-to-

ambient CO2 concentrations (Ci/Ca) for Heliotropium species pooled according to photosynthetic functional type. Data are median

values ± standard error and were taken from the curves depicted in Figure 2.1 except the Ci/Ca data which corresponds to the PWUE

data in Table 2.2. Data were compared using a nested one-way ANOVA. Pairwise comparisons were performed using Tukey’s HSD

test. C3 includes H. europaeum, H. tenellum, H. karwinskyi, H. calcicola, and H. fallax; C3-C4 includes H. convolvulaceum, H. racemosum, and H. greggii; C4 includes H. texanum. Letters indicate significant differences at P<0.05.

Species Slope value R2 -1 (mmol CO2 mol O2) C3 species Heliotropium europaeum 2.81 ± 0.07 0.99 Heliotropium karwinskyi 2.81 ± 0.07 0.78 Heliotropium tenellum 2.81 ± 0.07 0.95 a Combined C3 slope 2.70 ± 0.08 0.68

Weak C3-C4 intermediate Heliotropium procumbens 2.42 ± 0.07a,b 0.86

C3-C4 intermediate species Heliotropium convolvulaceum 2.00 ± 0.14 0.94 Heliotropium greggii 1.80 ± 0.08 0.77 Heliotropium racemosum 1.42 ± 0.39 0.86 b Combined C3-C4 slope 1.86 ± 0.17 0.64

C4 species Heliotropium texanum -0.05 ± 0.02c 0.02

Table 2.4. The slopes of the response of the CO2 compensation point of photosynthesis to

-1 variation in O2 concentration below an O2 level of 210 mmol mol . Data from Figure 2.2.

Slopes were calculated for the O2 range corresponding to the initial linear portion of the response in the C3-C4 intermediate species. Mean ± SE, N=3 for individual species. The mean of the pooled slope values for the C3 and C3-C4 species are also shown. Superscripted letters indicate statistical differences between the pooled data for the C3 species, pooled data for the C3-C4 species, and data from H. procumbens and H. texanum. Statistical analysis used a nested one-way ANOVA with Tukey’s HSD test used for pairwise comparisons.

Species g370 g300 g200 g30 (= gmax) H. europaeum 0.380 ± 0.071a 0.384 ± 0.08a 0.386 ± 0.08a 0.646 ± 0.11

H. calcicola 0.301 ± 0.021a 0.362 ± 0.02a 0.374 ± 0.03a 0.783 ± 0.03

H. fallax 0.345 ± 0.042a 0.351 ± 0.06a 0.364 ± 0.04a 0.762 ± 0.03

H. tenellum 0.420 ± 0.050a 0.424 ± 0.06a 0.444 ± 0.05a 0.820 ± 0.06

H. karwinskyi 0.315 ± 0.077a 0.320 ± 0.10a 0.327 ± 0.11a 0.784 ± 0.16

H. procumbens 0.438 ± 0.036a 0.509 ± 0.03a 0.533 ± 0.03a 0.710 ± 0.04

H. convolvulaceum 0.314 ± 0.114a 0.425 ± 0.12a 0.446 ± 0.12a 0.653 ± 0.14

H. racemosum 0.291 ± 0.034a 0.293 ± 0.02a 0.293 ± 0.02a 0.629 ± 0.04

H. greggii 0.329 ± 0.028a 0.376 ± 0.02a 0.411 ± 0.03a 0.887 ± 0.10

H. texanum 0.315 ± 0.019a 0.341 ± 0.02a 0.345 ± 0.02a 0.984 ± 0.03

Table 2.5. Stomatal conductances measured at CO2 concentrations of 370, 300, 200 and 30

-1 μmol mol . Data were taken from A/Ci curves depicted in Figure 2.1. Sample sizes are the same as those listed in Table 2.2. Data were taken at leaf temperatures of 30°C, vapor pressure deficit of 18 mmol mol-1, and photosynthetic photon flux density of 1200 μmol m-2 s-1.

A 40

-1 30 s -2 m 2 20 H. tenellum H. calcicola H. fallax mol CO mol 10 H. convolvulaceum µ H. racemosum H. texanum 0 B 40

Netphotosynthesis rate, 20

H. europaeum 10 H. karwinskyi H. procumbens H. greggii 0 0 100 200 300 400 500 600 700 Intercellular CO concentration, µmol mol-1 2

Figure 2.1.

Figure 2.1 (A) and (B) Net CO2 assimilation rate as a function of intercellular CO2 concentration determined for the Heliotropium species in this study. Data are separated into two panels to improve clarity. All responses were measured at 30°C, a leaf-to-air vapor pressure deficit of 18 mmol mol-1 and photosynthetic photon flux density of 1200 μmol

-2 -1 photon m s . Arrows indicate measurements taken at the current CO2 level of 370 μmol mol-1. Mean ± SE, N = 2 to 6 as indicated in Table 2.2.

-1 160 H. europaeum H. tenellum 140 H. karwinskyi H. procumbens mol mol

µ 120 H. convolvulaceum H. racemosum H. greggii 100 H. texanum 80

40 compensation point, point, compensation

2 20

CO 0 0 100 200 300 400 500 -1 O2 concentration, mmol mol

Figure 2.2 The response of the CO2 compensation point to variation in O2 concentration for the Heliotropium species in this study except H. calcicola and H. fallax. CO2 compensation points were measured as the x-intercepts of A vs. Ci response curves at 30°C, VPD of 18 mmol mol-1 and PPFD of 1200 μmol photons m-2 s-1. Mean ± SE, N =3. See Table 2.4 for regression slopes.

Discussion

Gas exchange and carbon isotope data

Of the ten species examined, H. texanum is functionally C4, while H. europaeum, H. calcicola, H. fallax, H. tenellum, and H. karwinskyi are functionally C3. Heliotropium convolvulaceum, H. racemosum, and H. greggii are functionally C3-C4, as indicated by reduced Г, and the lower initial slope and curvilinear nature of the Г vs. pO2 response. These three species have Kranz-type anatomy as indicated by enlarged bundle sheath cells with increased chloroplast density (Frohlich 1978; Muhaidat 2007). They also localize glycine decarboxylase in the bundle sheath cells (Muhaidat et al. 2006; Muhaidat 2007), a characteristic common in all known intermediate species (Monson & Rawsthorne 2000).

Heliotropium procumbens may also be a C3-C4 intermediate species, although weakly so as indicated by its C3-like Г value. Muhaidat and co-workers observed it has enlarged bundle sheath cells, and its glycine decarboxylase is restricted to the bundle sheath tissue (Muhaidat et al. 2006; Muhaidat 2007). While a Г value of 48.5 µmol mol-1 observed in H. procumbens is within the range of C3 Г values, it is lower than Г in all the other C3 Heliotropium species examined, which is consistent with the operation of a weak C2 pathway. While it is clear that any C2 pathway activity in H. procumbens is of minor importance for carbon balance at

30°C, it may be more significant in extreme photorespiratory conditions associated with heat and drought events in its native habitat.

The values of Γ from H. convolvulaceum, H. racemosum and H. greggii are similar to values obtained for other putative Type I C3-C4 species. In Type I intermediates of Flaveria

(Ku et al. 1983; Ku et al. 1991), Moricandia (Brassicaceae) (Apel et al. 1997), Alternanthera

(Amaranthaceae) (Rajendrudu et al. 1986), Mollugo (Molluginaceae) (Sayre & Kennedy

1977), Salsola (Chenopodiaceae) (Voznesenskaya et al. 2001), and Parthenium (Asteraceae)

-1 (Moore et al. 1987), Γ is generally between 15 and 30 µmol mol . This stands in contrast to the known Type II species Flaveria floridana, F. palmeri, and F. brownii which express Γ values ≤10 µmol mol -1 (Ku et al. 1991). With the addition of the Г data from the

Heliotropium species, a clear pattern emerges where discrete clusters of intermediates, as characterized by Г, are apparent. The clusters are near the C3 compensation point of 50 µmol mol-1 in putative weak Type I intermediates, near 20 µmol mol-1 in fully expressed Type I intermediates, and ≤10 µmol mol-1 in type II intermediates. Notably lacking are Г estimates between 30 and 40 µmol mol-1 at 25°C to 32°C. These results indicate there may be only a few viable levels of intermediacy in the evolution of C4 plants. If true, C4 evolution would likely occur in jumps between the discrete stages, rather than in gradual, slight shifts between intermediate modes. Identifying the anatomical and molecular changes that facilitate evolutionary jumps may be useful in identifying ways to engineer the C4 pathway into C3 plants.

Associated with reduced Γ in the C3-C4 intermediates is an enhanced PWUE (Table

2.2). The PWUE in the C4 H. texanum was more than twice the mean value in the C3 species, while the C3-C4 intermediates H. convolvulaceum, H. racemosum and H. greggii had mean

PWUE values that were somewhat elevated from C3 values. C3-C4 Panicum species also have intermediate PWUE values (Brown & Simmons 1979), although Type I Flaveria intermediates have similar PWUE as C3 Flaveria species (Monson 1989). Increased PWUE would enhance performance in water-limited or saline environments and contribute to the success of C3-C4 intermediates at high temperatures and VPD. Our results show that the enhancement of PWUE in Heliotropium intermediates is not a result of lower stomatal

conductance when compared with C3 species. Rather, the higher PWUE of the intermediates is primarily a product of greater photosynthetic rate, particularly at lower CO2 levels.

Stomatal conductance did not differ between the species at 200, 300 and 370 μmol mol-1

CO2. Corresponding to these results, Ci/Ca values at ambient CO2 were lower in the C3-C4 intermediates relative to that observed in C3 Heliotropium species. In a similar comparison,

Huxman and Monson (2003) found that C3-C4 Flaveria species exhibited similar relative conductance values as the C3 Flaveria species. Our results and those of Huxman and

Monson (2003) indicate stomatal regulation has not fundamentally changed with the evolution of the C2 pathway in either Flaveria or Heliotropium. The rise of the more conservative stomatal regulation associated with C4 photosynthesis apparently occurred later in the evolutionary sequence within these two genera.

The response of Γ to O2 reflects the type and strength of intermediacy present. A reduction in the slope of Γ versus O2 reflects the operation of a stronger carbon-concentrating mechanism (Ku et al. 1991). Type I species such as Flaveria anomala, F. pubescens, and

Neurachne minor exhibit biphasic responses with Γ values intermediate between C3 and C4 levels at lower O2 levels (Apel & Maass 1981; Hattersley et al. 1986). Type II species such as F. floridana have a shallow response slope that approaches those observed in C4 plants

(Holaday, Lee & Chollet 1984; Ku et al. 1991; Sudderth et al. 2007). Heliotropium convolvulaceum, H. racemosum, and H. greggii show responses typical of Type I

-1 intermediates as the slope of the Γ vs. O2 response below 210 mmol mol is approximately a third less than was observed in C3 species, and there is an increase in the slope above 210 mmol mol-1, similar to Type I Flaveria intermediates (Ku et al. 1991; Dai, Ku & Edwards

1996). The increase in slope is thought to occur where the CO2 refixation capacity by

Rubisco in the bundle sheath is exceeded by photorespiratory activity (Monson &

Rawsthorne 2000).

13 All of the C3 and C4 Heliotropium species examined had δ C values that are typical of their respective photosynthetic types (Table 2.2). The values for the three strong intermediates are similar to mesophytic C3 species, which is consistent with Type I C3-C4 intermediacy (von Caemmerer 1989). In Type I intermediates, both the original fixation of

CO2 in the mesophyll and the refixation of photorespired CO2 in the bundle sheath occur via

Rubisco; hence, the leaf carbon isotope signature is dominated by Rubisco discrimination as it is in C3 plants (von Caemmerer & Hubick 1989).

Ecological considerations

The similarity in A at elevated CO2 between H. greggii and the C3 species H. karwinskyi and H. europaeum provides a clear description of the consequences of having the

C2 pathway in terrestrial plants. The rate of A at elevated Ci is similar between H. greggii and these two C3 species, as are the initial slopes. Because of lower Г, H. greggii has a larger value of A at lower and current CO2 levels than the C3 species. This reduction in Γ reflects the high degree of CO2 refixation that occurs in the relatively isolated bundle sheath compartment (Monson & Rawsthorne 2000).

The importance of the intermediates’ enhanced photosynthetic rates and higher water- use efficiencies to performance in natural environments becomes clear when considering these species’ native habitats in the context of recent geological time. Heliotropium racemosum is an annual plant of sandy, disturbed sites on the Texas coastal plain while H. convolvulaceum is an annual sand-dune specialist in arid regions of the American Southwest

59 and Mexico. Heliotropium greggii occupies open sites on sand, clay or gravel soils in the

Chihuahuan Desert area of New Mexico, Texas and Mexico (Correll & Johnston 1970). All three species are summer-active and receive substantial moisture from summer monsoons.

All routinely experience high air temperature (>40°C) during the peak of their growing season, and all generally occur in arid to semi-arid situations. As an example, on a Mojave

Desert sand dune near Las Vegas, Nevada, I observed August leaf temperatures in H. convolvulaceum exceeded 40°C between mid-morning and late afternoon, while soil temperatures exceeded 55°C (P. Vogan, H. Coiner & R. Sage, unpublished data). At 30°C, the temperature of our A/Ci measurements, the presence of C3-C4 intermediacy enhances A in

H. convolvulaceum, H. greggii and H. racemosum by 40% to 60% relative to the C3 species

-1 at atmospheric CO2 levels representative of the late Pleistocene epoch (200 µmol mol ). At

-1 early industrial era CO2 levels near 300 µmol mol , the enhancement of A in the C3-C4 intermediates is 20% to 40% greater than values in the two C3 species (Table 2.3).

Presumably, during similar periods of the past, further enhancement would have occurred in their natural environments where high temperatures and low humidities are common, and photorespiration would have been high. In a similar case, the C3-C4 intermediate Flaveria floridana maintains photosynthetic rates four times higher than a sympatric C3 species at 35°

-1 to 42°C at CO2 levels of 340 µmol mol (Monson & Jaeger 1991), further demonstrating the potential contributions of C3-C4 metabolism to carbon balance for plants in hot environments.

Conclusion

The identification of H. convolvulaceum, H. racemosum, H. greggii and possibly H. procumbens as C3-C4 intermediates provides a new system to examine the functional significance of traits that compensate for photorespiration. Our results show that these traits improve photosynthetic performance in low CO2 conditions equivalent to those of the recent past. The situation may soon change, however, for at higher than current CO2, there is no obvious advantage to the C2 pathway as indicated by the A/Ci responses of H. greggii and H. karwinskyi. The precise relevance of the C2 pathway to the question of C4 evolution in

Heliotropium is unclear at this time due to the lack of a detailed, species-level phylogeny for section Orthostachys. A recent phylogeny by Hilger and Diane (2003) indicates H. convolvulaceum and H. procumbens branch between C3 ancestors and derived C4 clades; however, the sampling of species within section Orthostachys was relatively low, thereby limiting resolution. Despite this limitation, their phylogeny supports the hypotheses that the

C2 pathway evolved prior to C4 photosynthesis in section Orthostachys and may be a key stage in C4 evolution. With the addition of phylogenetic data, Heliotropium can become an important new model in which to study the evolution of C4 photosynthesis.

Chapter Three - Water-use and nitrogen-use efficiency of C3-C4 intermediate species of Flaveria Juss. (Asteraceae)

Abstract

Plants using the C4 pathway of carbon metabolism are marked by greater photosynthetic water-use and nitrogen-use efficiencies (PWUE and PNUE, respectively) than

C3 species, but it is unclear to what extent this is the case in C3-C4 intermediate species. In this study, I examined the PWUE and PNUE of fourteen species of Flaveria Juss.

(Asteraceae), including two C3, three C4 and nine C3-C4 species, the latter containing a

14 gradient of C4-cycle activities (as determined by initial fixation of C into C-4 acids). I found that PWUE, PNUE, leaf Rubisco content and intercellular CO2 concentration (Ci) in air do not change gradually with C4-cycle activity. These traits were not significantly different between C3 species and C3-C4 species with less than 50% C4-cycle activity while intermediates with greater than 65% C4-cycle activity were not significantly different from plants with fully-expressed C4 photosynthesis. These results indicate that a gradual increase in C4-cycle activity has not resulted in a gradual change in PWUE, PNUE, intercellular CO2 concentration and leaf Rubisco content towards C4 levels in the intermediate species. Rather, these traits arose in a stepwise manner during the evolution of C4-like intermediates, which are contained in two different clades within Flaveria.

Introduction

The C4 photosynthetic pathway suppresses oxygenation of ribulose-1,5-bisphosphate

(RuBP) by concentrating CO2 around ribulose-1,5-bisphosphate carboxylase/oxygenase

(Rubisco), using a metabolic cycle that begins with the fixation of inorganic carbon via phosphoenolpyruvate carboxylase (PEPC). C3-C4 intermediate species also possess a

61 62

mechanism to reduce photorespiratory CO2 loss below C3 levels, using a pathway of carbon metabolism that is distinct from both C3 and C4 photosynthesis. In a typical C3 leaf, much of the CO2 released in photorespiration would diffuse out of the leaf, particularly in warm and arid habitats where rates of RuBP oxygenation are high (Rawsthorne 1992; Sage & Pearcy

2000). C3-C4 species minimize photorespiratory CO2 loss by compartmentalizing photorespiration such that the step in which CO2 is released (glycine decarboxylation) is restricted to bundle-sheath cells. Much of the CO2 that is released by glycine decarboxylation is recaptured by bundle-sheath chloroplasts before it can diffuse out of the leaf (Hylton et al. 1988; Rawsthorne et al. 1988). This results in higher rates of net CO2 assimilation at low CO2 and high temperature in C3-C4 species than in C3 species, and reduced CO2 compensation points of photosynthesis (Г; Schuster & Monson 1990; Monson

& Jaeger 1991; Vogan et al. 2007).

A universal feature of the C4 pathway is enhanced water- and nitrogen-use efficiency of photosynthesis. In all evolutionary lineages examined, the C4 species have photosynthetic water-use efficiencies (PWUE) that are typically 1.5 to 4 times greater than the C3 species under similar measurement conditions (Larcher 2003; Vogan et al. 2007; Kocacinar et al.

2008). Maximum nitrogen-use efficiency of photosynthesis (PNUE) in C4 plants is 50-100% greater than in closely related C3 plants from distinct evolutionary lineages (Brown 1978;

Monson 1989; Li 1993; Ripley et al. 2008). These differences in PWUE and PNUE have profound consequences for trophic dynamics and ecosystem function. For example, the higher PWUE of C4 plants allows for greater productivity on limited water and longer growing seasons in seasonally dry environments (Long 1999; Knapp & Medina 1999; Sage et al. 1999; Sage & Pearcy 2000). In arid regions, the high PWUE of C4 plants allows for

vegetation coverage that may otherwise be absent, and C4 grasses comprise the majority of warm-season grass species in semi-arid environments (Teeri & Stowe 1976; Sage et al. 1999;

Wan & Sage 2001). In dry, fire-prone habitats, the greater biomass production of C4 plants accelerates fuel accumulation rates, thereby promoting the intensity and frequency of wildland fire, which in turn promotes the grassland biome (Knapp & Medina 1999). In grasslands, high biomass productivity on limited amounts of nitrogen produces vegetation swards of low nitrogen content (Wedin & Tilman 1996). These are relatively difficult to digest and decompose, such that rates of biogeochemical cycling are reduced (Berendse et al.

1987; Ross et al. 2002). To efficiently consume C4 vegetation, herbivores required specialized digestive adaptations that helped drive faunal diversification through evolutionary time (Van Dyne et al. 1980; Phelan & Stinner 1992; Heckathorn et al. 1999). It is possible, therefore, to conclude that the greater PWUE and PNUE of C4 vegetation influenced the origin of the modern biosphere.

While it is clear that the C4 pathway enhances PWUE and PNUE, it is unclear which aspects of C4 photosynthesis are responsible for these enhancements, and when in the evolutionary sequence from C3 to C4 photosynthesis the enhancements occurred. On the one hand, enhanced PWUE and PNUE could occur in a steady progression from the C3 ancestors to the C4 progeny. Alternatively, there could be a sudden increase in PWUE or PNUE that is associated with the evolution of a particular trait of the C4 pathway. With respect to PWUE, support has been presented for both possibilities for the genus Flaveria (Asteraceae).

Flaveria is a widely studied lineage that has three C3 species, five C4 species, and 14 species with varying degrees of intermediate characteristics between full C3 and C4 photosynthesis.

In Flaveria, PWUE has either been interpreted as increasing steadily along the sequence

from full C3 to full C4 species (Huxman & Monson 2003), or increasing in a step-wise manner late in the evolutionary sequence, when C4-like species evolve (Kocacinar et al.

2008). Other examinations of Flaveria indicate C3-C4 intermediates have C3 values of

PWUE under current ambient CO2 concentrations (Monson 1989; Apel 1994). PNUE changes along a C3 to C4 evolutionary transition are difficult to evaluate from the literature, as PNUE variation in C4 lineages has not been comprehensively examined. In Flaveria, two

C3-C4 species have C3-like PNUE values, while a third, Flaveria ramossissima, exhibits a value in between a C3 and C4 species (Monson 1989).

In order to address evolutionary hypotheses, it is important to have a well-resolved phylogeny of the lineage in question. Until recently, a lack of well-resolved phylogenies has hampered analysis of trait evolution from C3 ancestors to C4 progeny. In Flaveria, a well- resolved, species-level phylogeny was produced in 2005 by McKown et al. This phylogeny demonstrates two distinct lineages of C4 and C4-like photosynthesis evolving from common

C3 ancestors (Fig. 3.8). In clade “A”, four C4-species are present, along with two C4-like species, and one C3-C4 intermediate species, F. ramossissima (McKown et al. 2005).

Flaveria ramossissima branches in an evolutionarily intermediate position between the C3 and C4 species. In clade “B” a dozen C3-C4 intermediate species are present, but there are no full C4 species and only one C4-like species, Flaveria brownii. C4-like Flaveria species have

14 CO2 compensation points (Γ) and C-fixation values that are generally indistinguishable from C4 species; these plants are recognized as intermediates only because they lack complete localization of Rubisco to the bundle-sheath cells as occurs in C4 plants (Reed &

Chollet 1985; Moore et al. 1989). Of the C3-C4 intermediate species in Flaveria, six lack significant enhancement of PEPC activity and show Γ values of 20 to 30 µmol mol-1 (Ku et

al. 1991). These species are termed Type I C3-C4 intermediates (Edwards & Ku 1987).

-1 Three C3-C4 Flaveria species have low Γ (9-10 µmol mol ) and significant increases in

PEPC activity, indicating a weak to moderate C4 cycle is active (Ku et al. 1991). These are termed Type II intermediates (Edwards & Ku 1987). Both Type I and Type II intermediates in Flaveria show C3-like PWUE values based on instantaneous measurements of net CO2 assimilation rate (A) per transpiration rate (E) (Apel 1994; Kocacinar et al. 2008). PNUE differences between Type I and Type II intermediates are unclear, as only three C3-C4 intermediates – including two Type I and one Type II Flaveria species - have been evaluated

(Monson 1989).

Most measurements of PWUE are simple ratios of photosynthetic rate versus transpiration (A/E or A/gs at constant vapor pressure difference; for example, Kocacinar et al.

2008), and most measures of PNUE are A per unit leaf nitrogen content. Simple ratios do not account for non-photosynthetic costs of water and nitrogen in a leaf. For example, cuticular transpiration could skew PWUE values, while nitrogen associated with walls, storage pools, and non-photosynthetic tissues could skew PNUE values. The most robust measure of

PWUE is the slope of the relationship between A and leaf conductance, gs (Wong et al. 1979;

Schulze & Hall 1982), because it directly shows the sensitivity of A to an incremental unit of water lost as the stomata open. Likewise, the best assessment of PNUE is the slope of A versus leaf nitrogen per unit area (Na), which is less sensitive to background levels of N and shows the increment of A per unit of invested N. Few measurements of the slope of A versus gs and A versus Na are available. Monson (1989) evaluated PWUE and PNUE in Flaveria as the slope of A versus gs, and A versus Na, respectively, but only on three C3-C4 intermediate species: F. floridana, F. pubescens and F. ramossissima.

In this study, I evaluated A versus gs, and A versus Na in 14 Flaveria species selected to represent the most important branches in the Flaveria phylogeny (Table 3.1; Fig. 3.8).

Species selected include two C3 species, four Type I intermediates, two Type II intermediates, all three C4-like species, and three full C4 species (Table 3.1). My objective was to evaluate where in the evolutionary sequence from C3 to C4 photosynthesis the increases in PWUE and PNUE occur. To assess the role of the C4 cycle in boosting PWUE and PNUE, I compared PWUE and PNUE estimates with the degree of C4-cycle activity

14 previously determined for each species as CO2 incorporation into C-4 acids (Table 3.1;

Rumpho et al. 1984, Monson et al. 1986, Moore et al. 1987, Chastain & Chollet 1989). I also expressed A as a function of Rubisco content. In C4 plants, an important advantage of the C4 metabolic cycle is an increase the efficiency of Rubisco use, because Rubisco operates near CO2 saturation and has low oxygenase activity (Sage 2001). Increased Rubisco-use efficiency (RBUE) is the main driver of enhanced PNUE (Hikosaka & Shigeno 2009). In this particular study, I observed a non-linear response of A versus Na in most of the Flaveria species, indicating there were many non-photosynthetic pools of N in leaves of plants receiving fertilizer. In these instances, the analysis of Rubisco-use efficiency proved critical in evaluating the change in PNUE along the C3 to C4 transition in Flaveria.

Materials & methods

Seed sources and growth conditions

Seeds of F. kochiana and F. angustifolia were contributed by Dr. Erika Sudderth,

Harvard University (Sudderth et al. 2007). Seeds of all other species were contributed by Dr.

Gerald Edwards, Washington State University and Dr. Peter Westhoff, University of

Dusseldorf, and were originally drawn from the collection of Dr. A.M. Powell, Sul Ross

State University (Powell 1978; Monson et al. 1988).

Plants were grown from cuttings of at least three different individuals of each species and raised in growth chambers (Model GC-20, Enconair Ecological Chambers, Winnipeg,

MB, Canada) from January through August 2008. Day/night growth temperatures were

30/22°C, photoperiod was fourteen hours, and mean irradiance ± SE was 554 ± 9 µmol photons m-2 s-1. Plants were grown in an 8 L of a mixture containing equal parts sand, Pro-

Mix potting medium (Premier Horticulture, Inc., Quakertown, PA, USA) and sterilized topsoil. Plants were rotated weekly within each chamber to prevent edge effects. Plants were also trimmed weekly to minimize self-shading and slow soil water depletion. Species and treatments were distributed evenly among four growth chambers during the experiment to reduce the possibility of introducing error from environmental differences between chambers.

Plants were fertilized three times weekly with a modified Hoagland solution (Sage &

Pearcy 1987a). To acquire a range of photosynthetic capacities, plants were fertilized with

- 2- one of six N concentrations (16, 8, 4, 2, 1 and 0.5 mmol). The ions Cl and SO4 replaced

- the NO3 that was removed to produce reduced-nitrogen fertilizers.

Photosynthetic gas exchange and leaf nitrogen analysis

Measurements of photosynthetic gas exchange were conducted using a LI-6400

Portable Photosynthesis System (Li-Cor, Inc., Lincoln, NE, USA) on the youngest fully expanded leaf. Light in the leaf cuvette was raised gradually to 2000 µmol photons m-2 s-1 before taking steady-state measurements of net CO2 assimilation rate, leaf conductance to

H2O, and intercellular CO2 concentration. Leaf temperature was 30°C and cuvette CO2 concentration was set to 350 ppmv. Leaf-to-air vapor pressure difference (VPD) was set to

1.5 kPa.

After gas exchange was measured, two 2.56 cm2 leaf discs were removed from the leaf. One was dried overnight at 70°C, weighed, and assayed for total leaf nitrogen content using a Costech elemental combustion system (Costech Analytical Technologies, Inc.,

Valencia, CA, USA). The second was frozen in liquid N2 immediately after sampling and stored in a -80°C freezer. These discs were later assayed for total Rubisco content using 14C-

CABP binding method, with rabbit-raised antibody precipitation of Rubisco-14C-CABP complexes (Sage et al. 1993). Chlorophyll was assayed spectrophotometrically following extraction with acetone after Porra (2006).

Instantaneous PWUE (A/E) was calculated as PWUE = (Ca - Ci)/1.6ΔW where Ca is the ambient partial pressure of CO2 and ΔW is the leaf-to-air vapor pressure difference,

-1 which was 15 mmol H2O mol air for all measurements (Farquhar & Sharkey 1982).

Statistical analysis

Data for graphs of A v. g, A v. Na and A v. leaf Rubisco content were analyzed using regression analysis to determine if significant relationships existed between the variables, both by species and by functional type. These regressions were then compared using an analysis of co-variance (ANCOVA) to determine if significant differences existed between functional types in slope and y-intercept values at P < 0.05 (SAS statistical software, Cary,

NC, USA). In the case of A v. leaf Rubisco content, I plotted the regression lines through the origin as a lack of data at low leaf Rubisco levels introduced high amounts of variance into

69 the y-intercept values. This is also consistent with the physiology of photosynthesis, as net

CO2 assimilation rates will be near zero when leaf Rubisco content is nil (Bjorkman 1981;

Field & Mooney 1986). Values of A/E, A/Na, leaf Rubisco content and Ci were calculated and compared with an analysis of variance (ANOVA) on ranks to determine if there were significant differences between functional types, and pairwise comparisons were conducted using Dunn’s test to analyze differences between species.

Results

Effect of functional type on water- and nitrogen-use efficiency

The response of A v. g was assessed by species and functional type. These data were fitted with linear regressions (Figs. 3.1A & 3.1B), and there was a significant linear relationship found for each functional type and species (Table 3.2). The regressions were analyzed with an ANCOVA and significant differences in response slope and y-intercept existed between functional types (F = 182.2, df = 4, P < 0.0001). C3 species and the Type I and II C3-C4 intermediates did not differ significantly from each other in either the A v. g response slope or y-intercept (Table 3.2). The slopes and y-intercepts of the C4 species and

C4-like intermediates were significantly different from the C3 and Type I and II responses (P

< 0.05), but were not significantly different from each other. Values for the x-intercept of the A v. g responses were roughly similar between all functional types (-0.207 to -0.320) with no consistent differences between photosynthetic types. Also, instantaneous values of

PWUE (A/E) were compared using an ANOVA on ranks (Table 3.2). Functional types were significantly different (H = 166.592, df = 4, P < 0.001) and pairwise comparisons showed that C3 species and both Type I and II intermediates were not significantly different from

each other, but had significantly lower instantaneous PWUE than C4 and C4-like plants (P <

0.05). C4 and C4-like species were not significantly different from each other. Therefore, both intrinsic PWUE and instantaneous PWUE are significantly greater in the C4 and C4-like species than the other functional types, and PWUE is not increased in the Type I and II intermediates over the C3 levels.

Nitrogen-use efficiency followed a similar pattern. A nonlinear, quadratic-type relationship was found between A and Na in all functional types (Table 3.3, Fig. 3.2A &

3.2B). Regressions were analyzed using an ANCOVA and functional types were found to be significantly different (F = 7.52, df = 4, P = 0.0066). The linear portion of the response was measured for slope by plotting a linear regression through data points below N-saturation (≤

120 mmol m-2) (Table 3.3). Initial slopes were not significantly different between functional types (P > 0.05) (Table 3.3). In contrast, values of A/Na were compared using an ANOVA on ranks (Table 3.3), and functional types were found to be significantly different (H =

103.7, df = 4, P < 0.001). Pairwise comparisons indicate that C3 plants and Type I and II intermediates were not significantly different from each other, but had significantly lower instantaneous PNUE than C4 and C4-like plants (P < 0.05). C4 and C4-like plants did not have significantly different values of A/Na.

Because the increased instantaneous PNUE of C4 species is generally attributed to lower leaf Rubisco content (Sage 2001), and because the initial response slope of A v. Na did not differ significantly between functional types, I evaluated Rubisco-use efficiency (the slope of A v. leaf Rubisco content) in each species and functional type. Plots were fitted with linear regressions (Fig. 3.3A, 3.3B, Table 3.4) and a significant linear relationship was found in each species and functional type across the full range of leaf N. Due to a lack of data

71 points at low leaf Rubisco values, regressions were plotted through the origin. This is consistent with the observation that at zero leaf Rubisco content, net CO2 assimilation rate will be at or near zero (Bjorkman 1981; Field & Mooney 1986; Evans 1989). Using an

ANCOVA, functional types were found to differ significantly in their responses (F = 40.7, df

= 4, P < 0.0001). The slope of the C4 and C4-like responses did not differ significantly, but were both significantly greater than those of the C3 species and the Type I and II intermediates, which did not differ significantly from each other. Leaf Rubisco content per unit area was significantly different between functional types using an ANOVA on ranks (H

= 106.267, df = 4, P < 0.001). Pairwise comparisons using Dunn’s test showed no significant difference in leaf Rubisco content between C3 species and Type I and II intermediates, but did show a significant difference between these functional types and the C4 and C4-like individuals (P < 0.05). There was no significant difference between the C4 and C4-like species.

To assess if relative allocation of nitrogen to photosynthesis changes across leaf N levels, values of Rubisco N and chlorophyll N were divided by the corresponding total leaf

N. Also, relative allocation to thylakoid N content was calculated by assuming thylakoid N contains 55 mol N mol chl-1 (Evans 1983). Thylakoid N includes N allocated to chl-protein complexes, the electron transport chain and coupling factors, which contain approximately

55 mol N mol-1 chl (Evans 1983; Sage et al. 1987). Relative allocation of N to Rubisco was constant across leaf N levels (Fig. 3.6), and was greater in C3 species and Type I and II intermediates (12-17% of total leaf N content) than in C4 and C4-like species (4-8%).

Relative allocation of N to thylakoid components decreased linearly as total leaf N increased

(Fig. 3.6), from 25-50% of total leaf N at 50 mmol m-2 total leaf N content to 10-20% when

72 total leaf N content was 200 mmol m-2. There were no significant differences in thylakoid N content between functional types.

Influence of C4-cycle activity

Values of intrinsic PWUE, RBUE, leaf Rubisco content, and intercellular CO2 concentration were measured for each species and plotted against the percentage of 14C these species fix into C-4 acids during an 8-10 s pulse (Fig. 3.4 & 3.5). These values do not change gradually as C4-cycle activity increases. Rather, PWUE and RBUE increase 2-2.5 fold and leaf Rubisco content and Ci decrease 35% and 50%, respectively, between 50 and

65% fixation of 14C to C-4 acids (Fig. 3.4A, 3.4B, 3.5A & 3.5B).

14 Species CO2 compensation % C fixed to Functional point ± SE C-4 acids ± type SE F. pringlei 62.0 ± 0.3 4.1 ± 1.9 C3

F. cronquistiii 60.4 ± 1.7 7.7 ± 2.5 C3

F. sonorensis 29.6 ± 1.0 3.0 ± 0.0 Type I, C3-C4

F. chloraefolia 29.0 ± 2.0 11.3 ± 2.7 Type I, C3-C4

F. angustifolia 24.1 ± 0.4 11.0 ± 0.0 Type I, C3-C4

F. pubescens 21.3 ± 1.2 24.9 ± 8.4 Type I, C3-C4

F. floridana 9.5 ± 2.0 36.1 ± 8.0 Type II, C3-C4

F. ramosissima 9.0 ± 1.7 43.0 ± 3.0 Type II, C3-C4

F. vaginata 3.0 ± 1.2 68.0 ± 0.0 C4-like, C3-C4

F. brownii 6.0 ± 1.3 68.5 ± 5.3 C4-like, C3-C4

F. palmeri 4.7 ± 0.3 75.5 ± 0.5 C4-like, C3-C4

F. trinervia 3.5 ± 0.4 81.4 ± 5.4 C4

F. bidentis 3.2 ± 0.3 72.0 ± 0.0 C4

F. kochiana 2.2 ± 0.6 N/A C4

Table 3.1 Biochemical characteristics and photosynthetic functional types of Flaveria species in this study. Data for CO2 compensation points from Ku et al. (1991) except F. kochiana from Sudderth et al. (2007). Data for %14C fixed to C-4 acids were pooled from

Rumpho et al. (1984), Monson et al. (1986), Moore et al. (1987), and Chastain & Chollet

(1989), and mean and standard error (SE) were calculated.

Species Functional type Instantaneous Response slope y-intercept R2 F P-value PWUE ± SE (µmol CO2 (µmol CO2 -1 -1 -2 -1 (mmol CO2 mol mmol H2O) m s ) H2O F. cronquistii C3 4.81 ± 0.20 36.4 8.6 0.87 107.3 <0.001 F. pringlei C3 4.37 ± 0.21 21.4 11.5 0.79 62.4 <0.001 a a a All C3 4.58 ± 0.15 29.4 9.4 0.70 83.5 <0.001 F. angustifolia Type I C3-C4 4.64 ± 0.36 24.4 12.1 0.93 143.1 <0.001 F. chloraefolia Type I C3-C4 4.84 ± 0.25 35.6 6.3 0.84 132.8 <0.001 F. pubescens Type I C3-C4 4.62 ± 0.11 28.9 13.7 0.64 36.3 <0.001 F. sonorensis Type I C3-C4 4.11 ± 0.11 31.8 11.0 0.78 72.4 <0.001 a a a All Type I C3-C4 4.63 ± 0.11 36.7 7.6 0.88 596.8 <0.001 F. floridana Type II C3-C4 4.09 ± 0.16 27.9 12.5 0.49 13.2 0.003 F. ramosissima Type II C3-C4 4.49 ± 0.11 42.9 6.2 0.91 133.6 <0.001 a a a All Type II C3-C4 4.29 ± 0.11 39.4 9.9 0.71 72.4 <0.001 F. brownii C4-like 7.48 ± 0.21 77.2 10.7 0.89 92.0 <0.001 F. palmeri C4-like 9.01 ± 0.26 89.1 12.3 0.80 44.4 <0.001 F. vaginata C4-like 7.13 ± 0.19 62.3 11.3 0.72 38.8 <0.001 b b b All C4-like 7.80 ± 0.18 57.6 15.7 0.68 89.1 <0.001 F. bidentis C4 7.85 ± 0.26 60.4 12.7 0.62 25.0 <0.001 F. kochiana C4 9.16 ± 0.25 70.7 14.2 0.90 146.8 <0.001 F. trinervia C4 8.98 ± 0.26 68.5 12.2 0.82 74.3 <0.001 b b b All C4 8.69 ± 0.16 63.9 13.8 0.72 130.8 <0.001

Table 3.2 Linear regression analysis of data for the relationship of net CO2 assimilation rate versus leaf conductance for Flaveria species and functional types in this study. Letters indicate significant differences between functional types at P < 0.05.

Species Functional type Instantaneous Initial slope y-intercept R2 F P-value PNUE (µmol CO2 (µmol CO2 (µmol CO2 mmol-1 N s-1) mmol-1 N s-1) m-2 s-1) F. cronquistii C3 0.169 ± 0.010 0.235 -10.7 0.60 11.3 0.001 F. pringlei C3 0.169 ± 0.013 0.087 6.9 0.19 1.9 0.190 a a a All C3 0.169 ± 0.008 0.164 -2.3 0.40 11.3 <0.001 F. angustfolia Type I C3-C4 0.188 ± 0.014 0.270 -31.5 0.34 2.5 0.129 F. chloraefolia Type I C3-C4 0.205 ± 0.016 0.199 -4.7 0.38 7.4 0.003 F. pubescens Type I C3-C4 0.225 ± 0.012 0.061 23.2 0.18 2.0 0.161 F. sonorensis Type I C3-C4 0.223 ± 0.011 0.310 -28.3 0.50 9.4 0.001 a a a All Type I C3-C4 0.212 ± 0.007 0.228 -4.4 0.51 41.7 <0.001 F. floridana Type II C3-C4 0.249 ± 0.020 0.087 14.6 0.28 2.5 0.123 F. ramosissima Type II C3-C4 0.267 ± 0.053 0.164 6.4 0.41 4.2 0.042 a a a All Type II C3-C4 0.257 ± 0.029 0.188 6.6 0.51 14.3 <0.001 F. brownii C4-like 0.273 ± 0.024 0.223 -7.1 0.55 6.1 0.018 F. palmeri C4-like 0.477 ± 0.063 0.156 22.3 0.33 2.4 0.139 F. vaginata C4-like 0.377 ± 0.013 0.365 -20.6 0.78 24.7 <0.001 b a a All C4-like 0.376 ± 0.023 0.245 -0.3 0.41 12.5 <0.001 F. bidentis C4 0.349 ± 0.027 0.198 7.8 0.76 22.6 <0.001 F. kochiana C4 0.245 ± 0.015 0.214 1.4 0.67 16.2 <0.001 F. trinervia C4 0.443 ± 0.026 0.315 -0.6 0.76 23.9 <0.001 b a b All C4 0.344 ± 0.017 0.159 14.3 0.64 45.2 <0.001

Table 3.3 Quadratic regression analysis of data for the relationship between net CO2 assimilation rate and leaf nitrogen content for

Flaveria species and functional types in this study. Letters indicate significant differences between functional types at P < 0.05.

Species Functional type Response slope R2 F P-value -1 (µmol CO2 g Rubisco s-1) F. cronquistii C3 16.4 0.76 40.2 <0.001 F. pringlei C3 10.8 0.25 5.8 0.032 a All C3 13.0 0.35 6.1 0.020 F. angustifolia Type I C3-C4 15.2 0.50 12.1 0.005 F. chloraefolia Type I C3-C4 15.5 0.66 24.8 <0.001 F. pubescens Type I C3-C4 16.3 0.63 28.4 <0.001 F. sonorensis Type I C3-C4 16.2 0.74 46.0 <0.001 a All Type I C3-C4 16.0 0.78 221.0 <0.001 F. floridana Type II C3-C4 18.1 0.51 12.6 0.004 F. ramosissima Type II C3-C4 27.7 0.78 31.1 <0.001 a All Type II C3-C4 20.2 0.54 26.6 <0.001 F. brownii C4-like 61.4 0.52 10.7 0.008 F. palmeri C4-like 59.1 0.25 3.0 0.021 F. vaginata C4-like 82.7 0.32 5.7 0.034 b All C4-like 53.8 0.26 3.9 0.018 F. bidentis C4 69.8 0.49 15.2 0.001 F. kochiana C4 47.6 0.28 6.7 0.019 F. trinervia C4 81.2 0.22 3.6 0.014 b All C4 57.2 0.29 18.6 <0.001

Table 3.4 Linear regression analysis of data for the relationship of net CO2 assimilation rate versus leaf Rubisco content for Flaveria species and functional types in this study. Letters indicate significant differences between functional types at P < 0.05.

Figure 3.1

Figure 3.1 Response of leaf conductance to photosynthetic capacity for Flaveria functional types in this study, organized as (A) raw data by species and (B) as linear regressions by functional type. C3 species are represented by “3,” C4 species by “4,” and Type I, Type II and C4-like C3-C4 species by I, II and III, respectively.

Figure 3.2

Figure 3.2 Response of net CO2 assimilation rate to leaf nitrogen content for Flaveria functional types in this study, organized as (A) raw data and (B) quadratic regressions.

Figure 3.3

Figure 3.3 The relationship between net CO2 assimilation rate and leaf Rubisco content for

Flaveria functional types in this study, organized by (A) species as raw data and (B) functional types as linear regressions. C3 species are represented by “3,” C4 species by “4,” and Type I, Type II and C4-like C3-C4 species by I, II and III, respectively.

Figure 3.4 The relationship between (A) intrinsic water-use efficiency (the slope of A v. g) and (B) intercellular CO2 concentration versus C4-cycle activity for Flaveria species in this study. Values for initial fixation of 14C to C-4 acids in a pulse-chase assay are from Table

3.1.

Figure 3.5 The relationship between (A) Rubisco-use efficiency (the slope of the response of A v. leaf Rubisco content) and (B) leaf Rubisco content versus C4-cycle activity for

Flaveria species in this study. Values for initial fixation of 14C to C-4 acids in a pulse-chase assay are from Table 3.1.

Figure 3.6 (A) Rubisco content expressed as a percentage of leaf nitrogen versus total leaf

N for Flaveria functional types in this study. (B) Thylakoid N content expressed as a percentage of leaf nitrogen versus total leaf N for Flaveria functional types in this study.

Figure 3.7 Response of net CO2 assimilation rate to intercellular CO2 concentration for

Flaveria pringlei (C3) at 2% and 21% O2, F. ramosissima (Type II C3-C4) and F. bidentis

(C4). Photosynthetic rates are normalized as the proportion of the maximum rate for each curve. The supply function is calculated after Farquhar & Sharkey (1982). The solid line represents F. bidentis, the long dashes F. pringlei at 21% O2, the short dashes F. pringlei at

2% O2 and the dots F. ramosissima.

Figure 3.8 Phylogeny of Flaveria from McKown et al. 2005. Reprinted with permission from American Journal of Botany.

Discussion

Water-Use Efficiency

Photosynthetic WUE was not significantly greater in Type I and II intermediates than in C3 species, indicating that the establishment and partial engagement of a C4-cycle does not by itself enhance PWUE. The slope of the A v. g response is over twice as great in C4 as in

C3 species (Table 3.2), demonstrating greater PWUE in the C4 species, as has been documented previously (Wong et al. 1979; Sage & Pearcy 1987b; Monson 1989; Vogan et al. 2007; Kocacinar et al. 2008). Response slopes of the Type I and II intermediate species are not significantly different from C3 plants, while the C4-like intermediates are not significantly different from the C4 species (Table 3.2, Fig. 3.1A & 3.1B). These patterns mirror those observed from A/g comparisons by Kocacinar et al. (2008) and are present in both clades of Flaveria where C4 or C4-like photosynthesis evolved. This step-wise, parallel rise in PWUE at the stage where C4-like photosynthesis evolves is a characteristic pattern of

C4 evolution within Flaveria.

The rise in PWUE is associated with a rise in the initial fixation of CO2 into C-4 acids above 50% of total fixation products (Fig. 3.4B). PWUE values are similar in all Flaveria species with less than 50% C4-cycle activity and PWUE increases rapidly to C4 levels in intermediates with greater than 65% C4-cycle activity. The evolutionary modifications that lead to the enhancement of fixation by the C4-cycle in F. brownii, F. palmeri and F. vaginata

13 are also associated with the reduction in carbon isotope discrimination (δ C) from C3 to near

C4 values, a substantial increase in carboxylation efficiency, and the establishment of high levels of PPDK activity (Ku et al. 1983; Bauwe 1984; Monson et al. 1988; Monson &

Rawsthorne 2000).

Thus, the greater C4-cycle activity present in C4-like Flaveria intermediates cannot fully explain the significantly greater PWUE observed in these species, as no effect of C4- cycle activity was found on PWUE in Type II intermediates. In addition to the amount of

CO2 fixed to C-4 products by PEPC, the localization of photosynthetic enzymes within the leaf may be an important determinant of the strength and efficiency of the limited C4 cycle present in certain C3-C4 Flaveria species, influencing carboxylation efficiency and, consequently, PWUE. In C4 leaves, PEPC activity is generally restricted to mesophyll cells, while the activities of decarboxylating enzymes, such as NADP-ME, are restricted to bundle- sheath cells (Reed & Chollet 1985; Moore et al. 1989). This partitioning produces a system in which carbon can be efficiently transferred from the C4 to C3 cycle. However, if these enzymes are not fully restricted to the different cell types, some amount of futile C4 cycling would occur in the mesophyll in which CO2 is fixed by PEPC and subsequently decarboxylated in the mesophyll, greatly reducing the efficiency of carbon transfer into the

PCR cycle from fully-C4 levels (Monson et al. 1986; Edwards & Ku 1987; Cheng et al.

1988; Ku et al. 1991; Casati et al. 1999; Monson & Rawsthorne 2000). Incomplete partitioning of photosynthetic enzymes is present in at least one Type II Flaveria species, F. ramosissima, and this may partly account for the C3 levels of PWUE observed in this species. In F. ramosissima, PEPC activity is roughly two times greater in mesophyll (M) than bundle-sheath (BS) protoplasts, and activity of the decarboxylating enzyme NADP-ME is about twice as great in BS as M (Moore et al. 1988). There is also no significant difference in Rubisco activity between BS and M in F. ramosissima. In contrast, the C4-like species F. brownii, F. palmeri and F. vaginata display a much sharper gradient of enzyme activities between BS and M. Mesophyll protoplasts exhibit 4-7 times greater PEPC activity

92 than BS, and NADP-ME activity is 6-12 times greater in BS than M (Cheng et al. 1988;

Moore et al. 1989). Also, Rubisco activity is 3-8 times greater in BS than M in these species.

The relatively more homogenous distribution of photosynthetic enzymes in F. ramosissima may account for the observation that carboxylation efficiency is also unchanged from C3 levels in this species (Ku et al. 1983). This may in turn account in part for the C3-like

PWUE observed in the Type II Flaveria intermediates, as a stronger partitioning of C4-cycle enzymes would result in higher carboxylation efficiency and, consequently, greater rates of net CO2 assimilation for a given amount of leaf transpiration. Taken together, these results indicate that a substantial C4 pump, characterized by at least 65% initial fixation of CO2 by

PEPC and strong partitioning of photosynthetic enzymes between bundle-sheath and mesophyll is necessary before there are significant increases in PWUE from C3 levels.

The pattern of PWUE observed in Flaveria intermediates also has implications for the evolution of the stomatal response to intercellular CO2 concentration. In this experiment and others, stomatal conductance declines directly with mesophyll photosynthetic capacity such that Ci is constant or nearly constant across the range of photosynthetic capacities (Fig.

3.1A & 3.1B; Wong et al. 1979; Wong et al. 1985; Sage & Pearcy 1987b; Monson 1989). It is thought that this response is regulated by stomata via a “CO2 feedback loop” in which intercellular CO2 concentration is relayed to guard cells using an as yet unidentified signal from the mesophyll (Farquhar et al. 1978; Farquhar & Sharkey 1982; Mott 2009). Huxman

& Monson (2003) found that when leaves are exposed to changing light levels, which leads to changes in Ci, the Ci-signalling component of the resulting stomatal response is stronger in

C4 than in C3 species. It was also relatively more important in the Type II species F. floridana than in C3 species, but not in the Type I species F. chloraefolia and F. sonorensis

(Huxman & Monson 2003). This alteration in stomatal mechanism in Type II intermediates has apparently not resulted in greater PWUE as may be expected, as no significant increase in PWUE was observed in F. floridana either by us or by Monson (1989). This may be because the trend observed by Huxman & Monson (2003) towards a stronger Ci feedback response in F. floridana was detected above normal operating Ci, but not at or below it, which is where our measurements were taken.

Another means of assessing stomatal maintenance of Ci is through measurement of the “supply function” of the A/Ci response. The supply function indexes the resistance to

CO2 diffusion through the stomatal pore, and a greater absolute value of its slope indicates lower resistance to diffusion (Farquhar & Sharkey 1982; Sage & Kubien 2003). This is useful in determining whether a decrease in Ci, such as that observed in C4 leaves, is simply a product of greater photosynthetic CO2 assimilation or if stomatal resistance to CO2 diffusion has increased as well, potentially indicating a fundamental change in stomatal regulation of

Ci. For example, in leaves of the C3 species Flaveria pringlei measured at 2% O2, there is a marked increase in A relative to 21% O2 and a corresponding decrease in Ci due to greater rates of photosynthetic CO2 consumption; however, the resistance to CO2 diffusion into the leaf has not changed as indicated by a constant slope in the supply function between the two oxygen concentrations (Fig. 3.7). This is consistent with an unchanged stomatal set point to

Ci. Similarly, the Type II C3-C4 species F. ramosissima exhibits no change in Ci and no change in the slope of the supply function from the C3 F. pringlei (Fig. 3.7). By contrast, the

C4 species F. bidentis exhibits a reduced slope of the supply function (Fig. 3.7), indicating that enhancement of photosynthetic rate only partially explains the decrease in Ci observed in this species. A fundamental change in stomatal regulation appears to have occurred in F.

bidentis such that stomata are regulated to maintain a lower mesophyll CO2 concentration than in the C3 F. pringlei. Increases in carboxylation efficiency alone cannot explain the increase in PWUE occurring in C4-like and C4 Flaveria species. It is also apparent from measurements of F. ramosissima that such a shift is not present in the Type II intermediate. I conclude that some fundamental change in the Ci-signalling pathway likely accompanied the rapid shifts in PWUE and Ci that occurred in the transition from Type II to C4-like intermediacy in Flaveria.

Nitrogen-Use Efficiency

The response of A to Na was curvilinear for all functional types (Fig. 3.2A & 3.2B), indicating that storage of N in non-photosynthetic pools is occurring in Flaveria leaves at leaf N contents above 150 mmol m-2. High N storage was also indicated by slow reductions in leaf N content following imposition of low N treatments (data not shown). Values of A were higher in C4 and C4-like species at a given Na, producing significantly greater instantaneous PNUE in these functional types (Table 3.3). However, the slope of the A v. Na response was not significantly different between C3 and C4 species due to large amounts of variance in the data sets (Table 3.3). Consequently, Rubisco-use efficiency (RBUE) was used in this study as the primary metric of nitrogen-use. This is because leaf Rubisco content is the primary driver of differences in PNUE between ecologically similar C3 and C4 species

(Sage 2001; Hikosaka & Shigeno 2009). The slope of the response of A v. leaf Rubisco content was significantly greater in C4 than in C3 species (Table 3.4, Fig. 3.3A). The slope of the C4-like response was not significantly different from C4, but was significantly greater than that of the Type I and II intermediate species, which were not significantly different

from the C3 plants. This result mirrors that of PWUE in that RBUE does not change significantly until a substantial C4 pump has been established in the C4-like species. The only possible exception to this trend is the Type II species F. ramosissima which has somewhat lower leaf Rubisco levels and somewhat higher RBUE than the Type I and II species. This is consistent with the higher levels of C4 cycle activity in this species and its reduced sensitivity to O2 inhibition (Ku et al. 1983; Monson et al. 1986; Ku et al. 1991), though the magnitude of increased RBUE is still well below the C4 and C4-like levels (Table

3.4).

The increase in instantaneous PNUE values towards C4 levels is associated with decreases in leaf Rubisco content (Fig. 3.5B), but not in leaf chlorophyll levels, which comprise only about 1-3% of total leaf N and were not significantly different between functional types (data not shown). Leaf Rubisco content does not appear to have evolved gradually as carbon metabolism became more C4-like in the Type I and II intermediates.

Rather, reduced expression of leaf Rubisco content apparently evolved rapidly as carbon metabolism became nearly C4-like in F. brownii, F. palmeri, and F. vaginata. These species originated in two separate clades within Flaveria (McKown et al. 2005), and it is apparent that the C4 and C4-like species from these clades have converged on similar values of leaf

Rubisco content.

Relative allocation of N to thylakoid components of photosynthesis (chlorophyll,

LHC and electron-transport proteins) decreases as total leaf N increases in all functional types, and relative allocation to Rubisco remains constant (Fig. 3.6). Increased nitrogen storage at high leaf N in some form unrelated to photosynthetic components likely accounts for a greater proportion of total leaf N at high leaf N values and accounts for the decline in

96 relative allocation to thylakoid components. This is consistent with the observation that A does not increase above leaf N of 150 mmol m-2.

Conclusions

The substantial shifts in PWUE, RBUE and instantaneous PNUE marking the transition from Type II to C4-like intermediacy in Flaveria demonstrate the importance of establishing an efficient C4 cycle for shifts in these ecophysiological traits. Increased engagement of PEPC in Type II intermediates does not result in greater PWUE or RBUE, consistent with the hypothesis that PEPC primarily serves in a CO2-scavenging role in the mesophyll of these species. PEP carboxylase does not become the dominant pathway for initial CO2 fixation until an efficient and well-compartmentalized C4 cycle is present. At that point, leaf Rubisco level and intercellular CO2 concentration drop substantially, greatly increasing PNUE and PWUE in C4-like intermediates. These results point to improved carbon economy as the primary benefit of the initial stages of C3-C4 intermediacy under ambient conditions, rather than changes in resource-use efficiency.

Chapter Four – Photosynthetic performance and acclimation of C3, C4 and C3-C4

species from three eudicot genera grown at low CO2 concentrations

Abstract

Variations in the atmospheric concentration of carbon dioxide exert strong influence on the productivity and fitness of plants using the C3 pathway of photosynthesis, and numerous studies have focused on the disparate performance of C3 and C4 plants under low

CO2 conditions. In this experiment, I tested for the first time the acclimation responses and photosynthetic performance of C3-C4 species, in concert with measurements of C3 and C4 species, in low CO2, high temperature conditions. I grew C3, C4 and C3-C4 species from

-1 three eudicot genera at CO2 concentrations of 380 and 180 µmol CO2 mol air and day/night temperatures of 37/29°C and assessed their photosynthetic performance and acclimation responses under these conditions. In terms of acclimation responses, there was no effect of growth CO2 concentration on nitrogen allocation to photosynthetic components such as

Rubisco and electron transport in any species. Similarly, there was no effect of growth CO2 concentration on stomatal responses to intercellular CO2 concentration or relative stomatal limitation of photosynthesis, indicating that there was no stomatal acclimation to low CO2 conditions. All photosynthetic types were largely constrained in their ability to acclimate to low CO2, and this is particularly important in the case of C3 species. During past periods of low CO2, such as the glacial periods of the Pleistocene, they would not be able to acclimate as a means of promoting greater photosynthetic efficiency in a low CO2 environment, potentially increasing selection pressure for ways to make carbon assimilation more efficient.

The primary differences exhibited between the C3 and C3-C4 photosynthetic types were that

97 98

C3-C4 species had approximately 50% greater rates of net CO2 assimilation than the C3 species measured at the growth conditions of 180 µmol mol-1 and 37°C, and the thermal optimum of photosynthesis was less sensitive to CO2 concentration in C3-C4 species.

Enhanced carbon balance likely favored C3-C4 photosynthesis during periods of low atmospheric CO2 concentration and enhanced plant fitness. This was likely a key development in the evolution of the C4 photosynthetic pathway.

Introduction

Since the mid-Cenozoic Era (35 Ma), atmospheric CO2 concentrations have

-1 -1 fluctuated between 180 and 300 µmol CO2 mol air (hereafter µmol mol ), not including anthropogenic CO2 emissions since the Industrial Revolution (Pagani et al. 1999; Pearson &

Palmer 2000; Royer et al. 2001). The decline in atmospheric CO2 concentration from much higher levels (> 1000 µmol mol-1) prior to 30 Ma (Tipple & Pagani 2007) had pronounced negative effects on the growth and, ultimately, fitness of plants using the C3 pathway of carbon assimilation. Low CO2 levels have the dual effect of reducing substrate availability for photosynthesis while simultaneously increasing the frequency of the competing oxygenase reaction of ribulose 1,5-bisphosphate carboxylase/oxygenase (Rubisco) and the subsequent release of previously-fixed CO2 in the photorespiratory pathway. This reduction in the efficiency of C3 carbon assimilation is amplified at high temperatures by a relative increase in Rubisco affinity for O2 relative to CO2 and a proportionally greater decrease in

CO2 than O2 concentration in solution (Jordan & Ogren 1984). Illustrating the ill effects of low CO2 and high temperature on C3 photosynthetic species, laboratory studies have shown

-1 that CO2 levels at or below 200 µmol mol lead to a roughly 50% reduction in net CO2

99 assimilation rates from 380 µmol mol-1 (Sage & Coleman 2001) and, correspondingly, over a

50% reduction in total biomass in C3 species such as oat, wild mustard, mesquite, Abutilon theophrasti, bean and tobacco when grown at daytime temperatures of 19-30°C (Polley et al.

1992; Johnson et al. 1993; Dippery et al. 1995; Cowling & Sage 1998; Ward et al. 1999;

Campbell et al. 2005). At temperatures above 35°C, reductions in total biomass exceed 75% in bean plants compared to a cool, ambient CO2 treatment (Cowling & Sage 1998). This directly influences C3 plant fitness as flower and seed production decline significantly at or below 200 µmol mol-1 compared to current ambient levels (Dippery et al. 1995; Campbell et al. 2005).

-1 When CO2 levels were below 200 µmol mol , such as during the glacial periods of the Pleistocene epoch (1.8-0.01 Ma), a selective pressure likely favored mechanisms to improve photosynthetic efficiency (Ehleringer et al. 1991; Cowling 2001). It is believed that the C4 pathway, which first arose 25-32 Ma (Christin et al. 2008; Vincentini et al. 2008), evolved in response to conditions promoting high rates of RuBP oxygenation such as low atmospheric CO2 concentration, high temperature, high light and aridity (Ehleringer et al.

-1 1991; Sage 2004). At a CO2 concentration of 380 µmol mol , the C4 pathway of carbon fixation has greater quantum yield (amount of CO2 assimilated per quantum of light absorbed) than the C3 pathway at temperatures above 22-24°C (Ehleringer et al. 1997); at

Pleistocene CO2 concentrations, this crossover point is around 15-20°C (Ehleringer et al.

1997). These modelled differences in quantum yield correspond roughly to the temperature threshold at which C4 plants are favored in the modern biosphere (Sage & Pearcy 2000).

Several studies have addressed the differential performance of C3 and C4 photosynthetic species under low CO2 and/or high temperature conditions, and while C3

100

species experience substantial, adverse effects on productivity, the effects on net CO2 assimilation rate and total biomass yield in C4 species are much less if not absent (Johnson et al. 1993; Dippery et al. 1995; Tissue et al. 1995; Ward et al. 1999; Ward et al. 2008). While these studies are informative as to the performance of these two photosynthetic types under atmospheres of the previous 35 Ma, no study has addressed the performance of C3-C4 intermediate species grown at low CO2. In at least two genera, Flaveria (Asteraceae) and

Alternanthera (Amaranthaceae), there is evidence to suggest that these species represent an intermediate state in the evolution of the C4 pathway (McKown et al. 2005; Sanchez-del Pino

2009). Because of the proposed role of low CO2 in the evolution of the C4 pathway, an evaluation of the performance of C3-C4 intermediates in low CO2 conditions is important for understanding the potential significance of C3-C4 traits in the evolution of C4 photosynthesis.

C3-C4 photosynthesis functions to capture more photorespired CO2 than would be recovered in a C3 leaf, thereby improving plant carbon balance (Ehleringer & Monson 1993).

This reduction in photorespiratory CO2 loss is manifested as greater rates of net CO2 assimilation at low intercellular CO2 concentrations (Ci) and lower CO2 compensation points

(Γ) than in C3 species (Edwards & Ku 1987; Ku et al. 1991). While several laboratory studies have compared the leaf-level photosynthetic performance of C3 and C3-C4 species in the range of 25-30°C, few have addressed performance at higher leaf temperatures (> 30°C) that are likely common among C3-C4 eudicots growing in the field (Monson & Jaeger 1991;

P. Vogan, unpublished data, Fig. 4.14). Many C3-C4 plants are summer-active species growing in warm, high light environments of the tropics and sub-tropics (Lundell 1966;

Correll & Johnston 1970; Frohlich 1978; Powell 1978; McKown et al. 2005).

101

While improved carbon economy is believed to be an important benefit of the C3-C4 pathway at low CO2 and high temperature, these growth conditions may have effects on plant performance and survival beyond carbon relations alone; they may also entail alterations in plant development and resource-use and partitioning. For example, tobacco plants grown at or below 200 µmol mol-1 remain longer in the seedling stage (leaf area < 1 cm2) when they are more vulnerable to stochastic fluctuations in temperature or water availability and, consistently, they experience much higher rates of seedling mortality (Campbell et al. 2005).

Also, low CO2 conditions exacerbate abiotic stresses such as low water and nutrient availability in C3 plants. As CO2 concentration is reduced, stomatal conductance will increase to facilitate greater CO2 diffusion into the leaf (Anderson et al. 2001; Maherali et al.

2002); however, this entails the trade-off of greater rates of transpirational water loss, and in warm and arid environments, leaves will either lose proportionally more water at low CO2 to maintain a given rate of net CO2 assimilation, or they will reduce stomatal conductance substantially, imposing low Ci levels to conserve water (Sage & Coleman 2001).

Correspondingly, it has been shown that growth at subambient CO2 results in significantly more negative xylem water potentials in C3 plants (Polley et al. 2002) and reduced photosynthetic water-use efficiency (Polley et al. 1993a,b; Anderson et al. 2001). High CO2 concentrations have the opposite effect and can partially alleviate moderate drought stress

(Hsiao & Jackson 2000). With regards to nitrogen relations, C3 species grown under low

CO2 conditions are hypothesized to increase nitrogen allocation to Rubisco relative to non- limiting processes such as electron transport (Campbell et al. 1988; Sage & Reid 1992; Sage

& Coleman 2001) and to increase investment in aboveground growth relative to root growth as a means of compensating for low CO2 availability (Dippery et al. 1995). These

102 acclimation responses, while serving to improve whole-plant carbon balance, diminish photosynthetic nitrogen-use efficiency as relatively more leaf N is required to maintain a given rate of net CO2 assimilation (Anderson et al. 2001) while, at the same time, reduced root growth inhibits the plant’s ability to acquire an adequate amount of this critical nutrient

(Dippery et al. 1995). These acclimation responses may be attenuated in C3-C4 species relative to C3 species if the increased recovery of photorespired CO2 in C3-C4 leaves reduces the stress imposed by conditions promoting high rates of RuBP oxygenation. As a result, C3-

C4 species may be able to reduce relative allocation to Rubisco and other N-intensive photosynthetic components compared to C3 plants at low CO2. They may also conserve water by exhibiting lower stomatal conductances than C3 plants without experiencing as substantial a reduction in net CO2 assimilation rate.

In this study, I grew plants under low and current atmospheric CO2 concentrations

(180 and 380 µmol mol-1, respectively) and at high temperatures (37/29°C day/night) to determine the potential benefits of the C3-C4 pathway to plant carbon balance and resource- use efficiency under conditions proposed to favor C4 evolution. Closely-related C3, C4 and

C3-C4 species from each of three genera (Flaveria, Heliotropium (Boraginaceae) and

Alternanthera) were studied to determine if responses of the different photosynthetic types to these conditions were similar across multiple evolutionary lineages. I measured the response of net CO2 assimilation rate (A) to Ci at 30°C and 40°C and of A to leaf temperature (T) at low and ambient CO2. I also assessed the response of stomatal conductance (g) to Ci to determine if stomatal acclimation was similar between photosynthetic types. Finally, I measured total leaf nitrogen, leaf Rubisco content, and leaf chlorophyll content to determine

103

if biochemical allocation to photosynthetic processes was altered by growth under low CO2 concentrations.

Materials and Methods

Source material and growth conditions

Seeds of Flaveria were donated by Dr. Gerald Edwards, Washington State University and were originally from the collection of Dr. A.M. Powell, Sul Ross State University

(Powell 1978; Monson et al. 1988). Seeds of Alternanthera were donated by Dr. Peter

Westhoff, University of Dusseldorf. Seeds of Heliotropium were collected in the field by R.

Sage and P. Vogan during summers of 2004 and 2006 (Vogan et al. 2007). Plant specimens are vouchered and stored at the Royal Ontario Museum, Toronto, ON.

Plants were germinated from seed and raised in growth chambers (Model GC-20,

Enconair Ecological Chambers, Winnipeg, MB, Canada) from February to December, 2009.

Seeds were initially germinated under a light bank outside the growth chambers for 2 weeks prior to being placed in chambers. Day/night growth temperature was 37/29°C, photoperiod was fourteen hours, and mean irradiance ± SE was 561 ± 11 µmol photons m-2 s-

1. Plants were grown in 8 L pots containing a mixture of equal parts sand, Pro-Mix potting medium (Premier Horticulture, Inc., Quakertown, PA, USA) and sterilized topsoil. Plants were rotated weekly within each chamber to prevent edge effects and trimmed periodically to minimize self-shading and to slow soil water depletion. Plants were watered daily and fertilized three times weekly with a full-strength Hoagland solution.

Plants of each species were grown under one of two CO2 levels ± SE: 184 ± 5 and

381 ± 8 µmol mol-1. Each chamber was fitted with a PP Systems WMA-2 infrared gas

104

analyzer (PP Systems International Inc., Amesbury, MA) to monitor CO2 concentration. To maintain CO2 at the desired levels, the gas analyzers activated fans mounted onto a fibreglass enclosure placed inside the chambers which contained Sodabsorb CO2 Absorbent (Smiths

Medical International Ltd., Keene, NH). This chemical was replaced every 2-3 days, which was sufficient to maintain CO2 concentrations at the desired levels. After one set of measurements was taken on all plants, the treatments were switched between the two chambers and the experiment repeated to factor out any environmental effects of the individual growth chambers.

Photosynthetic gas exchange and leaf nitrogen analysis

Measurements of photosynthetic gas exchange were conducted using a LI-6400

Portable Photosynthesis System (Li-Cor, Inc., Lincoln, NE, USA) on the youngest fully expanded leaf. For photosynthetic responses to temperature (A/T), light in the leaf cuvette was gradually raised to 1500 µmol photons m-2 s-1 and then the temperature was lowered from 25°C to 15°C and then raised in 5°C increments to 45°C. Temperature control was maintained with a water bath attached to the LI-6400 sensor head via a 6400-88 Expanded

Temperature Control kit. Photosynthetic responses to CO2 (A/Ci) were logged at 30°C and

40°C during the course of the A/T measurements. Leaf-to-air vapor pressure difference fluctuated somewhat during the A/T measurements between 1.5 and 3.5 kPa, but was maintained at 2-2.7 kPa for A/Ci measurements. The photosynthetic thermal optimum was calculated by fitting a cubic regression to the A/T responses and determining the relative maximum of the function. Carboxylation efficiency and Γ were calculated from A/Ci graphs by plotting a regression through the linear portion of the response. Data for the response of g

105

v. Ci were taken from the A/Ci data sets, and relative stomatal limitation of photosynthesis was calculated as

1 - (A/Ao) x 100%

where A is the rate of net CO2 assimilation at a given Ci, and Ao is the rate of net CO2 assimilation if stomatal resistance to CO2 diffusion were zero (Farquhar & Sharkey 1982).

Valencia, CA, USA). The second was frozen in liquid N2 immediately after sampling and stored in a -80°C freezer. These discs were later assayed for total Rubisco content using the

14C-CABP binding method, with rabbit-raised antibody precipitation of Rubisco-14C-CABP complexes (Sage et al. 1993). Chlorophyll was assayed spectrophotometrically following extraction with acetone (Porra 2006). Nitrogen allocation to electron transport was calculated from values of leaf chlorophyll content based on a relationship of approximately

55 mol N allocated to electron transport for every mol of chlorophyll (Evans 1983; Sage et al. 1987). Nitrogen allocation to Rubisco was calculated based on a protein nitrogen content of 16% (Field & Mooney 1986).

Results

Photosynthetic responses to intercellular CO2

106

The A/Ci responses of C3, C4 and C3-C4 species were similar between plants grown at ambient and low CO2. The CO2 compensation point was unaffected by growth at low CO2 as were carboxylation efficiency and net CO2 assimilation rate at the highest measurement CO2 level (A800) (data not shown), indicating that photosynthetic capacity was not altered in plants grown at reduced CO2 levels. Elevated temperature (40°C) led to increases in Γ in both C3 and C3-C4 species; the Γ increases in C3 species were about 50% greater than those in C3-C4 species (Table 4.1). CO2 compensation points in C4 plants were not significantly affected by

-1 temperature. Net CO2 assimilation rates at growth CO2 (Ca = 380 and 180 µmol mol ) were marginally lower in both C3 and C3-C4 species at 40°C compared to 30°C, but the decrease in

A was approximately 30% greater in C3 than in C3-C4 species (Fig. 4.1-4.3; Table 4.1). This

-1 was especially apparent at 180 µmol mol as C3-C4 species exhibited photosynthetic rates roughly 67% greater than C3 species (Table 4.1). It is also of note that F. ramosissima, a C3-

C4 species with some C4-cycle activity, exhibited lower Γ than A. tenella and H. convolvulaceum, but there was no apparent increase in A at the growth CO2 level in this species compared to the other C3-C4 intermediates (Table 4.1).

Photosynthetic responses to temperature

-1 The photosynthetic thermal optimum (TOPT) at 380 µmol mol for the C4 species ranged from 34.0-42.3°C, for the C3 species was 30.4-32.1°C and for the C3-C4 species was

31.6-31.9°C (Table 4.2). The photosynthetic thermal optimum was significantly greater in the C4 species than in the other two types, but there was no significant difference between C3 and C3-C4 species at ambient CO2 (Table 4.2). When CO2 concentration was lowered to 180

-1 µmol mol , the thermal optimum of the C4 species was not significantly affected (Table 4.2;

107

Fig. 4.4-4.6), but it shifted considerably in the other two photosynthetic types. In the C3-C4 species, TOPT decreased to 29°C while in the C3 species, TOPT declined to about 25°C, a significantly lower value (Table 4.2; Fig. 4.4-4.6). Also, A at TOPT was significantly greater

-1 in C3-C4 species than in C3 species when measured at 180 µmol mol , though not as great as in the C4 species (Table 4.2). When compared at the temperature of the growth environment

(37°C), A in the C3-C4 species was not significantly different from C3 species at ambient

CO2, but at low CO2, A in C3-C4 species at 37°C was significantly higher (38%; Table 4.2,

Fig. 4.3-4.6). This phenomenon was not consistent across all three genera included in this experiment, as A in A. sessilis and A. tenella were not significantly different at 37°C.

Stomatal acclimation to low CO2

The responses of g v. Ci and Ci/Ca v. Ca were logged simultaneously with A/Ci measurements at 30°C and 40°C. The low CO2 treatment had no apparent effect on the g v.

Ci response of any functional type (Fig. 4.7-4.9), and values of gmax (that is, g at the lowest

Ci) were not significantly different between low- and ambient-grown plants within any functional type (data not shown). Stomatal conductance to H2O was similar across all three photosynthetic types (Table 4.4). The ratio of intercellular-to-ambient CO2 concentration was greater at low CO2 in C3 and C3-C4 species than at ambient CO2, but the difference in A

-1 between C3 and C3-C4 species of Flaveria and Heliotropium at 180 µmol mol resulted in significantly lower Ci/Ca in the C3-C4 species (Table 4.4). This is further reflected in the responses of Ci/Ca v. Ca as C3 and C3-C4 species exhibited similar values of Ci/Ca at 380

-1 µmol mol and above, but Ci/Ca was lower in C3-C4 species at CO2 concentrations below

300 µmol mol-1 (Fig. 4.10-4.12). Given the similar values of g for these two photosynthetic

108

types at these CO2 concentrations, the decline in Ci/Ca is primarily driven by increases in A at low CO2. C4 species exhibited lower Ci/Ca than C3 and C3-C4 species at nearly every CO2 concentration.

Relative stomatal limitation (RSL) of photosynthesis was calculated for each species at 40°C, and I found that RSL was significantly greater at low CO2 in C3 and C3-C4 species than at ambient, though there was no significant difference between the two photosynthetic types (Fig. 4.13). CO2 concentration had a small but statistically insignificant effect on RSL in C4 species. Relative stomatal limitation ranged between 20% and 23% in C3 and C3-C4 species at 380 µmol mol-1, while at 180 µmol mol-1, it was 30-34%. Photosynthetic water- use efficiency was also assessed in low- and ambient-grown plants at 40°C, and while C3 and

-1 C3-C4 species showed no significant differences at 380 µmol mol , PWUE was marginally higher in F. ramosissima and H. convolvulaceum than in F. robusta and H. calcicola at 180

µmol mol-1 (P = .07 and .08, respectively; Table 4.4). While this was not a statistically significant difference, the disparity in PWUE between C3 and C3-C4 species at low CO2 was far greater than at ambient CO2.

Leaf nitrogen allocation under low CO2

Growth at low CO2 effected no significant changes in total leaf N, relative N allocation to Rubisco and electron transport, leaf C:N ratios, or specific leaf mass (Table

4.3). Leaf Rubisco content was significantly lower and C:N was significantly higher in C4 species than in the other functional types, but there was no apparent biochemical acclimation to low CO2 of leaf N allocation in any photosynthetic type (Table 4.3). Photosynthetic nitrogen-use efficiency was assessed at 40°C at low and ambient cuvette CO2 concentrations

109

(Table 4.4). C4 species exhibited significantly greater PNUE than C3 and C3-C4 species at both CO2 concentrations, and there were no significant differences between C3 and C3-C4 species at ambient CO2 (Table 4.4). At low CO2, the difference in PNUE between C3 and

C3-C4 species was greater than at ambient CO2, and in two genera, Flaveria and

Heliotropium, this difference was statistically significant (Table 4.4). Given that leaf N levels and N allocation to Rubisco and electron transport were not different between these two functional types (Table 4.3), the trend in PNUE was driven by the greater rates of A in

-1 the C3-C4 species at 180 µmol mol .

110

Species Type Leaf T Carboxylation Γ A180 A380 A800 (°C) efficiency 30 0.110 ± 0.013 A 50.9 ± 1.1 A 9.8 ± 1.1 A 19.1 ± 1.8 A 28.3 ± 1.8 A F. robusta C 3 40 0.086 ± 0.012 A 77.1 ± 4.0 B 6.7 ± 0.4 B 15.0 ± 1.0 B 30.1 ± 2.7 A 30 0.086 ± 0.005 A 54.2 ± 1.8 A 6.9 ± 2.1 B 14.2 ± 1.4 B 24.7 ± 1.6 AB A. sessilis C 3 40 0.089 ± 0.009 A 76.4 ± 2.9 B 4.9 ± 3.1 B 9.1 ± 1.2 D 22.9 ± 1.8 AB 30 0.089 ± 0.015 A 52.2 ± 2.0 A 7.9 ± 0.9 AB 13.1 ± 1.2 B 21.7 ± 1.5 B H. calcicola C 3 40 0.087 ± 0.020 A 77.4 ± 0.8 B 4.9 ± 1.0 B 10.0 ± 0.9 D 20.1 ± 1.1 B 30 .095 ± 0.021A 52.4 ± 1.4A 8.2 ± 1.6AB 15.5 ± 1.5AB 24.9 ± 1.6AB All C A B B BD AB 3 40 .087 ± 0.018 77.0 ± 2.6 5.5 ± 1.8 11.4 ± 1.1 24.4 ± 2.0 30 0.124 ± 0.012 A 11.6 ± 1.6 D 13.7 ± 2.2 C 19.4 ± 1.6 A 25.9 ± 2.2 AB F. ramosissima C -C 3 4 40 0.116 ± 0.021 A 21.4 ± 2.5 E 11.4 ± 1.6 AC 15.8 ± 2.7 B 24.7 ± 2.1 AB 30 0.085 ± 0.008 A 18.1 ± 1.2E 9.8 ± 2.2 A 16.9 ± 2.3 AB 29.0 ± 0.3 A A. tenella C -C 3 4 40 0.086 ± 0.007 A 32.5 ± 2.8F 5.2 ± 0.9 B 12.8 ± 1.6 BD 24.3 ± 1.4 AB 30 0.126 ± 0.021 A 22.4 ± 2.1E 10.9 ± 1.7 A 22.0 ± 1.8 A 33.7 ± 4.1 A H. convolvulaceum C -C 3 4 40 0.130 ± 0.026 A 45.8 ± 6.8 AF 11.1 ± 1.6 AC 24.1 ± 2.7 AC 32.8 ± 3.8 A 30 .112 ± 0.017A 17.4 ± 1.5E 11.5 ± 2.0AC 19.4 ± 1.9A 29.5 ± 2.1A All C -C 3 4 40 .111 ± 0.020A 33.2 ± 3.4F 9.2 ± 1.2A 17.6 ± 2.4AB 27.3 ± 2.6AB 30 0.693 ± 0.087 B 2.8 ± 1.2 C 24.0 ± 3.6 D 27.8 ± 2.5 C 32.5 ± 4.1 A F. bidentis C 4 40 0.593 ± 0.121 B 3.0 ± 1.1 C 22.8 ± 2.1 D 25.2 ± 2.6 C 29.1 ± 3.6 A 30 0.666 ± 0.170 B 2.8 ± 0.6 C 22.3 ± 1.8 D 28.7 ± 3.9 C 28.8 ± 3.4 AB A. caracasana C 4 40 0.774 ± 0.126 B 3.3 ± 0.9 C 25.0 ± 4.4 D 29.2 ± 4.1 C 30.1 ± 3.3 A 30 1.58 ± 0.58 C 3.4 ± 1.3 C 29.7 ± 3.0 D 30.7 ± 3.2 C 31.0 ± 4.4 A H. texanum C 4 40 2.20 ± 0.55 C 5.3 ± 1.7 C 31.6 ± 3.1 D 32.1 ± 4.6 C 32.5 ± 5.1 A 30 .980 ± 0.463BC 3.0 ± 0.9C 25.3 ± 2.7D 29.1 ± 3.3C 30.8 ± 4.2AB All C 4 40 1.19 ± 0.421BC 3.9 ± 1.0C 26.5 ± 3.6D 28.8 ± 3.6C 30.6 ± 3.7AB

Table 4.1

111

Table 4.1 Carboxylation efficiencies, CO2 compensation points (Γ) and net CO2 assimilation rates at growth CO2 concentrations and at CO2 saturation for Flaveria, Alternanthera and

Heliotropium species in this study. Low- and ambient- grown plants were pooled for each value because A/Ci responses were not different between the two treatments. Measurements were taken at leaf temperatures of 30 and 40°C, photon flux density of 1500 µmol m-2 s-1 and leaf-to-air vapor pressure difference of 2-2.7 kPa. Carboxylation efficiency and Γ were calculated by plotting a linear regression through the lowest five Ci values of each A/Ci graph with carboxylation efficiency as the slope and Γ as the x-intercept. All values are means ± standard error. Different letters indicate significant differences at P < 0.05.

112

Species Type Measurement TOPT (°C) A at TOPT A at 37°C (µmol - -2 -1 CO2 (µmol CO2 m CO2 m s ) concentration 2 s-1) (µmol mol-1) 180 25.0 ± 0.8 A 9.6 ± 0.5 A 7.1 ± 0.4A F. robusta C 3 380 30.4 ± 0.4 B 19.9 ± 1.6 B 19.0 ± 1.1B 180 25.2 ± 0.1 A 10.5 ± 0.9 A 8.6 ± 1.1A A. sessilis C 3 380 32.1 ± 0.7 B 16.1 ± 1.2 B 15.2 ± 1.0B 180 24.3 ± 0.6 A 8.2 ± 1.0 A 5.1 ± 1.8A H. calcicola C 3 380 31.3 ± 0.3 B 13.7 ± 1.3 C 13.2 ± 1.1C 180 24.8 ± 0.7A 9.4 ± 0.8A 6.9 ± 1.2A All C 3 380 31.3 ± 0.4B 16.6 ± 1.4B 15.8 ± 1.1BC 180 29.0 ± 0.2 B 13.8 ± 1.1 C 11.2 ± 1.5C F. ramosissima C -C 3 4 380 31.6 ± 0.4 B 19.6 ± 1.8 B 16.9 ± 1.4B 180 30.6 ± 0.8 B 11.8 ± 0.4 AC 9.5 ± 0.7AC A. tenella C -C 3 4 380 31.6 ± 0.3 B 17.1 ± 1.7 B 15.6 ± 1.2BC 180 27.7 ± 0.4 AB 12.3 ± 0.5 AC 10.8 ± 0.4C H. convolvulaceum C -C 3 4 380 31.9 ± 0.2 B 24.3 ± 4.1 D 24.2 ± 2.7D 180 29.1 ± 0.5B 12.6 ± 0.6C 10.5 ± 1.0C All C -C 3 4 380 31.7 ± 0.3B 20.3 ± 2.2BD 18.9 ± 1.6BF 180 34.0 ± 0.5 C 25.0 ± 0.8 D 24.1 ± 1.0D F. bidentis C 4 380 35.9 ± 0.7 C 29.6 ± 2.3 D 28.0 ± 2.3D 180 35.6 ± 0.6 C 25.2 ± 3.7 D 24.7 ± 2.4D A. caracasana C 4 380 37.8 ± 1.0 C 29.0 ± 2.8 D 28.8 ± 2.6D 180 39.1 ± 0.3 C 26.8 ± 3.2 D 25.8 ± 2.9D H. texanum C 4 380 42.3 ± 1.5 C 39.1 ± 3.8 E 37.3 ± 3.3E 180 36.2 ± 0.5C 25.7 ± 2.8D 24.9 ± 2.1D All C 4 380 38.7 ± 1.2C 32.6 ± 3.0D 31.4 ± 2.8DE

Table 4.2 Data from responses of net CO2 assimilation rate (A) to leaf temperature (T) for

Flaveria, Alternanthera and Heliotropium species in this study. Data include photosynthetic

thermal optima (TOPT), A at TOPT and A at growth temperature of 37°C. Measurements were

taken at the plant’s respective growth CO2 concentration.

113

Species Type Growth Leaf N Leaf Rubisco Leaf %N %N C:N ratio Specific leaf -2 -2 -2 [CO2] (mmol m ) (mg m ) chlorophyll allocation to allocation to (by mass) mass (g m ) (µmol m-2) Rubisco electron transport

180 167 ± 20A 1358 ± 159 A 517 ± 23A 10.2 ± 0.90 A 15.3 ± 0.5 A 6.23 ± 0.57 A 52.2 ± 1.9 A F. robusta C AB A A A A A A 3 380 187 ± 10 1660 ± 117 424 ± 29 11.1 ± 2.93 16.3 ± 3.7 7.20 ± 0.25 43.2 ± 5.6 180 121 ± 18 A 1214 ± 108 A 484 ± 36A 9.98 ± 0.86 A 19.4 ± 2.1 AB 7.56 ± 0.31 A 41.8 ± 0.9 A A. sessilis C 3 380 113 ± 12 A 1150 ± 159 A 464 ± 44A 9.67 ± 1.57 A 17.5 ± 1.8 A 7.53 ± 0.24 A 40.5 ± 0.8 A 180 115 ± 13 A 1191 ± 107 A 350 ± 25C 11.3 ± 1.40 A 15.0 ± 0.9 A 8.02 ± 0.11 A 39.9 ± 2.2 A H. calcicola C 3 380 126 ± 16 A 1232 ± 104 A 393 ± 29AC 12.1 ± 1.68 A 18.6 ± 1.5 A 8.66 ± 0.27 A 43.6 ± 6.4 A 180 134 ± 16A 1254 ± 116A 450 ± 27A 10.5 ± 0.96A 16.6 ± 1.2A 7.27 ± 0.33A 44.6 ± 1.8A All C 3 380 142 ± 13A 1347 ± 128A 427 ± 33A 11.0 ± 1.97A 17.5 ± 2.0A 7.80 ± 0.25A 42.4 ± 4.2A 180 154 ± 11 A 1335 ± 162 A 510 ± 25A 10.2 ± 0.55 A 19.3 ± 1.6 AB 6.99 ± 0.48 A 43.3 ± 3.4 A F. ramosissima C -C 3 4 380 150 ± 11 A 1367 ± 70 A 474 ± 28A 10.1 ± 1.48 A 16.5 ± 0.7 A 6.50 ± 0.16 A 43.0 ± 4.7 A 180 160 ± 14 A 1277 ± 104 A 535 ± 21A 11.7 ± 2.60 A 21.5 ± 3.3 AB 7.98 ± 0.87 A 43.0 ± 3.4 A A. tenella C -C A A AB A AB A A 3 4 380 148 ± 19 1383 ± 180 582 ± 25 9.22 ± 0.60 21.0 ± 2.7 7.51 ± 0.69 45.8 ± 3.0 AB AC A A A A B H. 180 171 ± 10 2092 ± 428 543 ± 23 13.5 ± 1.05 14.6 ± 2.0 7.01 ± 0.44 58.4 ± 2.0 C -C convolvulaceum 3 4 380 209 ± 16 B 2428 ± 19 C 586 ± 50B 14.4 ± 3.89 A 22.2 ± 1.8 AB 8.19 ± 0.57 A 57.9 ± 1.0 B 180 162 ± 12A 1568 ± 286A 529 ± 23A 11.8 ± 1.25A 18.5 ± 2.2A 7.33 ± 0.59A 48.2 ± 3.0A All C -C 3 4 380 169 ± 16AB 1726 ± 75A 547 ± 34A 11.2 ± 1.87A 19.9 ± 1.9AB 7.40 ± 0.50A 48.9 ± 3.3A 180 127 ± 12 A 506 ± 31 B 568 ± 32AB 4.73 ± 0.57 B 23.4 ± 2.8 B 9.44 ± 0.43 B 46.2 ± 4.3 A F. bidentis C 4 380 136 ± 15 A 555 ± 46 B 597 ± 43B 4.72 ± 0.67 B 26.0 ± 1.7 B 9.35 ± 0.79 B 42.8 ± 4.2 A 180 140 ± 14 A 399 ± 55 B 637 ± 41B 3.55 ± 0.31 B 22.3 ± 2.3 B 8.51 ± 1.01AB 48.2 ± 2.0 A A. caracasana C 4 380 162 ± 14 A 496 ± 48 B 552 ± 28AB 3.31 ± 0.50 B 22.2 ± 4.2 B 8.73 ± 0.86AB 41.1 ± 1.5 A 180 120 ± 18 A 518 ± 43 B 497 ± 20A 3.86 ± 1.18 B 20.0 ± 5.5 AB 9.96 ± 0.94 B 52.0 ± 10.0A H. texanum C 4 380 170 ± 34 AB 492 ± 117 B 581 ± 39AB 5.01 ± 0.70 B 25.9 ± 4.2 B 10.52 ± 0.33B 51.6 ± 3.2 A 180 129 ± 15A 474 ± 42B 567 ± 30AB 4.05 ± 0.69B 21.9 ± 3.6B 9.30 ± 0.86B 48.8 ± 5.8A All C 4 380 156 ± 20A 678 ± 68B 577 ± 36AB 4.35 ± 0.63B 24.7 ± 3.3B 9.53 ± 0.71B 45.2 ± 3.0A

114

Table 4.3 Leaf tissue chemistry and nitrogen allocation patterns of Flaveria, Alternanthera and Heliotropium species in this study grown at CO2 concentrations of 180 and 380 µmol mol-1. All values are means ± standard error. Different letters indicate significant differences at P < 0.05.

115

Species Type [CO2] PWUE PNUE (µmol Stomatal Ci/Ca -1 (µmol CO2 CO2 mmol conductance -1 -1 -2 mol H2O) N s ) (mol H2O m s-1) 180 102 ± 20 A .030 ± .007A .310 ± .021 A .87 ± .01 A F. robusta C 3 380 264 ± 87 AB .088 ± .010B .132 ± .043 B .70 ± .01 B 180 101 ± 37 A .038 ± .006A .272 ± .045 A .89 ± .01A A. sessilis C 3 380 199 ± 42 B .091 ± .014B .089 ± .041 B .73 ± .01B 180 70 ± 34 A .036 ± .008A .302 ± .037 A .90 ± .01A H. calcicola C 3 380 142 ± 36 AB .102 ± .003B .101 ± .041 B .74 ± .01B 180 91 ± 28A .035 ± .007A .295 ± .036A .89 ± .01A All C 3 380 202 ± 51B .094 ± .009B .107 ± .042B .72 ± .02B 180 189 ± 73 A .065 ± .009C .333 ± .036 A .80 ± .01 C F. ramosissima C -C 3 4 380 244 ± 46 AB .105 ± .019B .166 ± .033 B .75 ± .01BC 180 116 ± 23 A .055 ± .012AC .241 ± .034 A .87 ± .02A A. tenella C -C 3 4 380 235 ± 12 B .081 ± .009B .120 ± .036 B .72 ± .01B 180 200 ± 50 AB .050 ± .009AC .363 ± .064 A .82 ± .02C H. convolvulaceum C -C 3 4 380 253 ± 33 B .157 ± .034D .198 ± .047 B .68 ± .01BD 180 168 ± 48AB .057 ± .010C .312 ± .042A .83 ± .01C All C -C 3 4 380 244 ± 32B .114 ± .021B .161 ± .038B .72 ± .03B 180 470 ± 145 C .182 ± .018D .360 ± .040 A .49 ± .02D F. bidentis C 4 380 517 ± 110 C .174 ± .016D .151 ± .026 B .52 ± .02DE 180 393 ± 54 C .167 ± .018D .375 ± .021 A .43 ± .01D A. caracasana C 4 380 451 ± 66 C .198 ± .007D .171 ± .030 B .59 ± .03E 180 418 ± 50 C .234 ± .073E .283 ± .051 A .31 ± .03F H. texanum C4 380 552 ± 115 C .240 ± .022E .128 ± .067 B .39 ± .03F 180 427 ± 78C .194 ± .039DE .339 ± .038A .41 ± .02DEF All C 4 380 507 ± 98C .204 ± .015DE .150 ± .035B .50 ± .03DEF

Table 4.4 Photosynthetic water-use and nitrogen-use efficiencies (PWUE and PNUE, respectively), stomatal conductances and intercellular-to-ambient CO2 ratios (Ci/Ca) of

Flaveria, Alternanthera and Heliotropium species in this study at cuvette CO2 concentrations

-1 ([CO2]) of 180 and 380 µmol mol . All values are taken at a leaf temperature of 40°C and are means ± standard error. Different letters indicate significant differences at P < 0.05.

116

Ambient-grown, 40°C

Ambient-grown, 30°C

Low-grown, 30°C Low-grown, 40°C

Figure 4.1 Response of net CO2 assimilation rate to intercellular CO2 concentration for ambient-grown (380 µmol mol-1) and low-grown (180 µmol mol-1) Alternanthera species in this study. Measurements were conducted at a photon flux density of 1500 µmol m-2 s-1, leaf-to-air vapor pressure difference of 2-2.7 kPa and leaf temperatures of 30°C and 40°C, as indicated.

117

Ambient-grown, 40°C

Ambient-grown, 30°C

Low-grown, 30°C Low-grown, 40°C

Figure 4.2 Response of net CO2 assimilation rate to intercellular CO2 concentration for ambient-grown (380 µmol mol-1) and low-grown (180 µmol mol-1) Flaveria species in this study. Measurements were conducted at a photon flux density of 1500 µmol m-2 s-1, leaf-to- air vapor pressure difference of 2-2.7 kPa and leaf temperatures of 30°C and 40°C, as indicated.

118

Ambient-grown, 40°C

Ambient-grown, 30°C

Low-grown, 30°C Low-grown, 40°C

Figure 4.3 Response of net CO2 assimilation rate to intercellular CO2 concentration for ambient-grown (380 µmol mol-1) and low-grown (180 µmol mol-1) species in this study.

Measurements were conducted at a photon flux density of 1500 µmol m-2 s-1, leaf-to-air vapor pressure difference of 2-2.7 kPa and leaf temperatures of 30°C and 40°C, as indicated.

119

A. caracasana (C4)

A. tenella (C3-C4)

A. sessilis (C3)

Figure 4.4

120

Figure 4.4 Response of net CO2 assimilation rate to leaf temperature for Alternanthera species in this study. Cuvette CO2 concentrations for these measurements were equal to growth CO2 concentration. Data sets were fit with cubic regressions.

121

F. bidentis (C4)

F. ramosissima (C3-C4)

F. robusta (C ) 3

Figure 4.5

122

Figure 4.5 Response of net CO2 assimilation rate to leaf temperature for Flaveria species in this study. Cuvette CO2 concentrations for these measurements were equal to growth CO2 concentration. Data sets were fit with cubic regressions.

123

H. texanum (C4)

H. convolvulaceum (C3-C4)

H. calcicola (C ) 3

Figure 4.6

124

Figure 4.6 Response of net CO2 assimilation rate to leaf temperature for Heliotropium species in this study. Cuvette CO2 concentrations for these measurements were equal to growth CO2 concentration. Data sets were fit with cubic regressions.

125

Ambient-grown, 30°C

Ambient-grown, 40°C

Low-grown, 30°C Low-grown, 40°C

Figure 4.7 Response of stomatal conductance to intercellular CO2 concentration at 30°C and

40°C for ambient-grown (380 µmol mol-1) and low-grown (180 µmol mol-1) Alternanthera species in this study.

126

Ambient-grown, 30°C

Ambient-grown, 40°C

Low-grown, 30°C Low -grown, 40°C

Figure 4.8 Response of stomatal conductance to intercellular CO2 concentration at 30°C and

40°C for ambient-grown (380 µmol mol-1) and low-grown (180 µmol mol-1) Flaveria species in this study.

127

Ambient -grown, 30°C

Ambient -grown, 40°C

Low -grown, 30°C Low -grown, 40°C

Figure 4.9 Response of stomatal conductance to intercellular CO2 concentration at 30°C and

40°C for ambient-grown (380 µmol mol-1) and low-grown (180 µmol mol-1) Heliotropium species in this study.

128

Ambient -grown, 30°C Ambient -grown, 40°C

Low -grown, 30°C Low -grown, 40°C

Figure 4.10 Response of intercellular-to-ambient CO2 concentration (Ci:Ca ratio) versus

-1 ambient CO2 concentration at 30°C and 40°C for ambient-grown (380 µmol mol ) and low- grown (180 µmol mol-1) Alternanthera species in this study.

129

Ambient -grown, 30°C Ambient -grown, 40°C

Low -grown, 30°C Low -grown, 40°C

Figure 4.11 Response of intercellular-to-ambient CO2 concentration (Ci:Ca ratio) versus

-1 ambient CO2 concentration at 30°C and 40°C for ambient-grown (380 µmol mol ) and low- grown (180 µmol mol-1) Flaveria species in this study.

130

Ambient -grown, 30°C

Ambient -grown, 40°C

Low -grown, 30°C Low -grown, 40°C

Figure 4.12 Response of intercellular-to-ambient CO2 concentration (Ci:Ca ratio) versus

-1 ambient CO2 concentration at 30°C and 40°C for ambient-grown (380 µmol mol ) and low- grown (180 µmol mol-1) Heliotropium species in this study.

131

* * * * * *

Figure 4.13 Relative stomatal limitation of photosynthesis at ambient and low CO2 concentrations. Asterisks indicate significant difference between treatments at P < 0.05.

132

Figure 4.14 Leaf temperatures of Heliotropium convolvulaceum recorded near Overton,

NV, USA from 28 Aug. - 2 Sept. 2006.

133

Discussion

Photosynthetic acclimation to low CO2

One of the hypothesized acclimation responses to low CO2 in C3 plants is some combination of increased leaf N allocation to Rubisco and/or a decrease in allocation to electron transport capacity, which may be accompanied by an increase in carboxylation efficiency (Campbell et al. 1988; Sage & Reid 1992; Sage & Coleman 2001). Rubisco capacity limits CO2 assimilation rate at low CO2, while the capacity to generate RuBP

(electron transport) is non-limiting (Farquhar & von Caemmerer 1982), and it is thought that leaves may exhibit changes in allocation patterns to bring the limiting and non-limiting processes closer to equilibrium and to maximize photosynthetic nitrogen-use efficiency

(Evans 1989; Sage & Coleman 2001). Such photosynthetic acclimation responses have been observed sporadically, including an increase in allocation to Rubisco relative to electron transport in Abutilon theophrasti (Tissue et al. 1995), Solanum dimidiatum (Anderson et al.

2001) and tobacco (Campbell et al. 2005), though the latter was observed in only one of the four subambient CO2 treatments used. Much more commonly, many low CO2 studies do not indicate any acclimation of Rubisco content or electron transport capacity in C3 leaves grown at low CO2 (Campbell et al. 1988; Thomas & Strain 1991; Sage & Reid 1992; Johnson et al.

1993; Campbell et al. 2005), and this null result was replicated in this experiment.

None of the species in this study showed significant photosynthetic acclimation to low CO2 as there was no evidence of greater allocation to Rubisco or lower allocation to electron transport in any species. This is based on both direct measurements of leaf Rubisco, chlorophyll and total nitrogen contents (Table 4.3), and by assessment of carboxylation efficiency (Table 4.1), the latter of which is directly influenced by Rubisco capacity to

134

assimilate CO2 under saturating light (Farquhar & von Caemmerer 1982; von Caemmerer

2000). Correspondingly, I observed no change in leaf C:N ratio as well (Table 4.3). I initially hypothesized that if changes in leaf N allocation would occur, they might be attenuated in C3-C4 relative to C3 species due to greater recovery of photorespired CO2. This hypothesis was clearly not supported by the data; there was no significant change in N allocation to photosynthetic processes in any functional type of any species when grown at low CO2. The frequently-observed inability of C3 plants to reallocate proportionally more leaf N to limiting processes demonstrates that such an acclimation response is not a reliable means of compensating for CO2 deficiency. This lack of photosynthetic plasticity at low

CO2 could have constrained the Pleistocene C3 plant’s ability to ameliorate the stress of low

CO2 during its lifetime and may have provided for greater selective pressure on ways to increase CO2 assimilation efficiency without increasing relative N requirement, as is the case in C3-C4 species.

Stomatal acclimation to low CO2

The responses of g v. Ci and Ci/Ca v. Ca were measured as an index of stomatal acclimation to low CO2. It has been observed in at least one C3 species, Solanum dimidiatum

(Solanaceae), that maximum stomatal conductance (measured at the lowest Ci) was significantly greater in low CO2-grown plants, and this increased g was consistent across the subambient range of CO2 concentrations (Maherali et al. 2002). This response may indicate that plants grown at low CO2 maintain higher g as a means of offsetting low CO2 availability.

Such a response was not observed in any species in this experiment, as the g v. Ci responses generally overlapped between the low- and ambient-grown plants (Fig. 4.7-4.9) and no

135

significant difference in gmax was observed between treatments (data not shown).

Correspondingly, the Ci/Ca ratio of low- and ambient-grown plants did not differ

-1 significantly at 180 or 380 µmol mol (data not shown) The responses of Ci/Ca v. Ca were

-1 similar between C3 and C3-C4 species at 300 µmol mol and above, but at subambient CO2 concentrations, Ci/Ca was significantly lower in C3-C4 species (Fig. 4.10-4.12; Table 4.4).

The absence of differences between the two photosynthetic types in the g v. Ci data indicate that the lower Ci/Ca values of the C3-C4 species are driven by inherently greater A at low CO2 rather than differences in stomatal acclimation. These results indicate low plasticity in stomatal conductance in response to reduced growth CO2, and the inability of stomata to compensate for low CO2 availability by increasing CO2 diffusion into the leaf, even in well- watered conditions.

In addition to this result, I calculated the relative stomatal limitation of photosynthesis and found that RSL was greater at 180 than at 380 µmol mol-1, a phenomenon common to all three functional types in all three genera in this study, but especially so in the

C3 and C3-C4 species (Fig. 4.13). I hypothesized initially that g and RSL may be lower in

C3-C4 relative to C3 species, but I did not observe a significant difference between these functional types; while there is greater limitation of photosynthesis by stomata at low CO2 generally (Fig. 4.13), there was no significant difference between limitation in C3 (33%) and

C3-C4 (30%) species.

Carbon balance of photosynthetic types under low CO2 and high temperature

While the A/Ci responses of all nine species were unaffected by growth at low CO2

(Fig. 4.1-4.3), there are substantial differences between photosynthetic types that have

136

important ramifications for plant carbon balance. The primary difference between the C3 and

C3-C4 photosynthetic types was a significantly lower Γ at 30°C and 40°C and substantially

-1 greater A at 180 µmol mol in the C3-C4 species, particularly at 40°C (Table 4.1). The differences in Γ between the C3 and C3-C4 types observed at 40°C are greater than at 30°C.

-1 At 40°C, the C3-C4 Γ is 44 µmol mol less than the C3 Γ, while at 30°C, Γ differs by 35

-1 µmol mol . As Γ differences explain much of the difference in A at low Ci, the change in the

Γ difference between the C3 and C3-C4 species reflects a greater significance of photorespiratory CO2 refixation in the C3-C4 species at higher temperature. These results are driven by the enhanced capacity of the C3-C4 species to recapture photorespired CO2, and they reflect those of previous studies of comparative gas exchange physiology in these species (Ku et al. 1983; Rajendrudu et al. 1986; Ku et al. 1991; Vogan et al. 2007).

Differences in A at low Ci are not driven by differences in carboxylation efficiency, as the values of C3 and C3-C4 species were not significantly different at 30°C and 40°C (Table 4.1;

Fig. 4.1-4.3).

Also of note is the lack of substantial differences in the A/Ci responses of H. convolvulaceum and A. tenella compared with F. ramosissima which, unlike the other two

14 C3-C4 species, has partial C4-cycle engagement as indicated by fixation of C to C-4 products (Monson et al. 1986). While Γ is significantly lower in F. ramosissima in agreement with reported values in these species (Rajendrudu et al. 1986; Ku et al. 1991;

Vogan et al. 2007), CE and A180 are similar between all three intermediates (Table 4.1).

Thus, the greater C4-cycle activity in F. ramosissima has only a minor impact on the A/Ci response relative to the other intermediates.

137

The A/T responses of the different photosynthetic types further illustrate the benefits of the C3-C4 pathway to plant carbon balance at low CO2 and high temperature. While TOPT and A at TOPT are not significantly different between C3 and C3-C4 species at ambient CO2

-1 (Table 4.2), the reduction of CO2 to 180 µmol mol resulted in a significantly greater decline in TOPT in C3 species (6.5°C) than in C3-C4 species (2.6°C). Net CO2 assimilation rate at the thermal optimum was significantly greater in C3-C4 than in C3 species (Table 4.2) and also declined by a larger proportion in C3 (43%) versus C3-C4 (38%) species. The significantly greater reduction in TOPT and A at TOPT in C3 species at low CO2 can be explained by higher rates of photorespiratory CO2 loss in the C3 species (Monson & Rawsthorne 2000). Greater ability to recapture photorespired CO2 would moderate the effects of lowering CO2

-1 concentration to 180 µmol mol on TOPT and A at TOPT and at 37°C in the C3-C4 species.

The greater ability to reassimilate photorespired CO2 in the C3-C4 species appears to buffer them against the detrimental effects of low CO2 and high temperature on net CO2 assimilation rate and, subsequently, whole-plant carbon balance.

The significantly greater A of the C3-C4 species at high temperature and low CO2 should have significantly enhanced plant fitness in warm, open environments during episodes of low atmospheric CO2 such as the glacial periods of the Pleistocene. The effects of net

CO2 assimilation rate on plant growth and fitness can be substantial. Reductions in A at subambient CO2, such as those observed in this experiment, have resulted in equivalent reductions in yield when measured in Abutilon theophrasti, Avena sativa, Brassica kaber,

Phaseolus vulgaris and Prosopis glandulosa (Polley et al. 1992; Johnson et al. 1993;

Dippery et al. 1995; Cowling & Sage 1998; Ward et al. 1999; Campbell et al. 2005). Such reductions in plant growth can impact reproductive output. Vegetative biomass exerts

138 enormous influence on fruit number and seed set, and reductions in plant growth can translate to significantly lower plant fitness (Solbrig 1981; Farris & Lechowicz 1990).

Several assessments have also been conducted on the fitness of plants with mutations that reduce photosynthetic rate. For example, a 25-30% reduction in light-saturated A in single- gene mutants of Amaranthus hybridus results in significantly lower seed set, seed size and seedling survival rates in field and greenhouse plots with light levels greater than 400 µmol photons m-2 s-1 (Arntz et al. 1998, 2000a, b). Arabidopsis thaliana mutants with reduced chlorophyll content have roughly 15% lower rates of light-saturated A and, correspondingly, experience a roughly 15% reduction in growth rate and 25% reduction in seed size (Janacek et al. 2009). While differences in A may influence fitness indirectly through growth rate and carbohydrate storage (Arntz et al. 1998), there is still a significant impact of A on plant fitness in high light environments, especially those characterized by low water availability

(Dudley 1996; Arntz et al. 2000a). This is particularly relevant for the C3 and C3-C4 species included in this experiment are native to high light, often arid environments. For example, in the case of Heliotropium convolvulaceum, I observed leaf temperatures in excess of 45°C at a site near Overton, NV, USA, in late August and early September, 2006 (Fig. 4.14).

Monson & Jaeger (1991) observed leaf temperatures in excess of 40°C in natural environments of the Type II intermediate Flaveria floridana in Florida during May and June.

While direct observations of leaf temperature are generally lacking in other species, the species grown in this experiment are all native to open, high light environments in the tropics and subtropics (Lundell 1966; Correll & Johnston 1970; Frohlich 1978; Powell 1978;

McKown et al. 2005) and may regularly experience leaf temperatures near or above 40°C for a substantial part of the growing season.

139

The final effect of differences in A between C3 and C3-C4 species at low CO2 and high temperature is a modest increase in PWUE and PNUE in C3-C4 species from C3 levels.

While PWUE and PNUE were similar between the two functional types at 380 µmol mol-1, corresponding to previous results (Chapter Three, this thesis), PWUE was marginally greater and PNUE was significantly greater in Flaveria and Heliotropium intermediates when measured at 180 µmol mol-1 (Table 4.4). Because stomatal conductances and leaf N levels were not significantly different between the two photosynthetic types, these differences in

PWUE and PNUE were driven entirely by higher A180 in the intermediates. This may further add to the advantages exhibited by the C3-C4 species and under low atmospheric CO2 concentrations by improving the plant’s ability to assimilate carbon and add biomass under low water and nutrient regimes (Farris & Lechowicz 1990; Dudley 1996; Arntz & Delph

2001).

Conclusions

The results of this experiment demonstrate improved carbon economy as the primary benefit of C3-C4 intermediacy under low CO2 and high temperature. There was no observed acclimation to low CO2 via leaf N allocation patterns or stomatal responses in any photosynthetic type, and the greater rates of net CO2 assimilation of C3-C4 intermediates at high temperature and low CO2 stands out as the major difference between photosynthetic types under the experimental conditions, resulting in lower Γ and greater TOPT and A at TOPT.

The low CO2 environments of the past 2 Ma, and the warm to hot environments to which these species are native, point to increased photorespiration as a pertinent stress in these conditions. The relatively greater carbon assimilation capacity of C3-C4 intermediates under

140 these conditions and the elevated PWUE and PNUE may have provided advantages in plant growth and reproduction and sustained these species in the warm, low latitude environments of the Pleistocene. In contrast, C3 species are mostly constrained in their ability to alleviate

CO2 starvation through photosynthetic or stomatal acclimation and would generally have been unable to compensate at the leaf level for the reduced carbon balance brought about by declining atmospheric CO2. This inability to acclimate to low CO2 could have provided the opportunity for natural selection to favor CO2 conservation mechanisms such as photorespiratory CO2 scavenging.

141

Chapter Five – Discussion

The C4 photosynthetic pathway comprises substantial changes to leaf anatomy and biochemistry and has a strong influence on the performance and distribution of different photosynthetic types in the modern biosphere. Temperature, light, water and salinity have markedly different impacts on the ecological distribution of C3 and C4 photosynthetic species, and the physiological variation between these types leads to C4 dominance in warm, high light grasslands of the tropics and sub-tropics (Sage & Pearcy 2000). It also has consequences for interactions between plants, herbivores and decomposers, as the altered lignin and C:N content of C4 tissues from C3 levels demands greater specialization among species consuming C4 vegetation (Wedin & Tilman 1990; Ehleringer & Monson 1993; Kemp et al. 1994; Murphy et al. 2002; Garibaldi et al. 2007). The pathway also contributes to the relative importance of prominent C4 crops such as maize and sugarcane, which are among the most productive of commercially cultivated plant species in low latitudes (Brown 1999;

Samson et al. 2005). Also, the prevalence of the C4 pathway in taxonomically distinct lineages indicates that it is one of more interesting examples of convergence in evolutionary biology, and the complex nature of the C4 pathway can provide insights into the evolution of complex physiological traits (Keeley & Rundel 2003; Sage 2004).

The study of C4 plant biology is greatly enhanced by an understanding of C3-C4 intermediacy. The phylogenetic position of C3-C4 species of Flaveria and Alternanthera indicates that they likely represent an important intermediate state in the evolution of the C4 pathway (McKown et al. 2005; Sanchez-Del Pino 2009). Understanding the advantages of species with the C3-C4 pathway under conditions promoting high rates of photorespiration, particularly high temperature and low atmospheric CO2 concentration, is imperative for

141 142

identifying the selection pressures favoring C4 photosynthesis and the anatomical and physiological traits that may have been necessary precursors to the pathway’s evolution

(Ehleringer et al. 1991; Sage 2004; McKown & Dengler 2007). Studies of C3-C4 intermediacy are also relevant to experimentation attempting to introduce the C4 pathway into C3 crops from warm environments, such as rice, as a means of increasing photosynthetic efficiency and improving yields (Leegood 2002; Sheehy et al. 2007; Furbank et al. 2009).

Understanding how C4 photosynthesis was derived from C3 photosynthesis in a stepwise fashion in nature may provide a blueprint for the stepwise engineering of C4 traits into C3 crops. This thesis addresses the carbon balance and resource-use efficiencies of C3-C4 eudicots in warm environments to provide greater understanding of the relative advantages of this pathway under conditions promoting high rates of photorespiration and to develop alternative model systems for understanding C4 evolution apart from the widely-studied

Flaveria.

I.) C3-C4 intermediacy in Heliotropium and the development of new model systems to study C4 evolution

In the second chapter of this thesis, I determined evaluated the presence and extent of

C3-C4 intermediacy in four species of Heliotropium: H. convolvulaceum, H. racemosum H. greggii, and H. procumbens. Based on measurements of Γ and O2 sensitivity, the first three species should be classified as Type I intermediates, as their CO2 compensation points and the sensitivity of Γ to O2 do not differ from other eudicot Type I intermediates measured previously (Ku et al. 1983; Holaday et al. 1984; Rajendrudu et al. 1986; Ku et al. 1991) and do not approach the levels of Type II intermediates in Flaveria (Ku et al. 1991). This is

143 further supported by low levels of PEPC in these species that do not approach the levels of

Type II intermediates of Flaveria, although PEPC activity in H. greggii is somewhat greater than in the other intermediate species (Muhaidat 2007). An important future contribution to this system would be the measurement of 14C fixation to C-4 products as a definitive means of assessing the extent of C4-cycle activity present in these species. This study of gas exchange physiology also contributed an especially novel finding: the first characterization of a “weak” Type I C3-C4 intermediate, H. procumbens, based on its nearly C3-like Γ coupled with the characteristic C3-C4 restriction of glycine decarboxylase to the bundle-sheath, as determined by Muhaidat (2007). It is also of note that C3 species exhibiting some traits of

C3-C4 intermediacy, such as the enlarged bundle-sheath cells of H. tenellum and H. karwinskyi or the slight enhancement of GDC localization to the bundle-sheath interior in H. karwinskyi (Muhaidat 2007), are not significantly different in any aspect of gas exchange physiology from those with smaller bundle-sheath cells such as H. europaeum. C3

Heliotropium species that exhibit some C3-C4 traits show a similar Γ to the other C3

Heliotropium species surveyed, and there is no apparent reduction of photorespiration via enlarged bundle-sheath cells or slight enhancement of bundle-sheath GDC localization in these species.

These characterizations of photosynthetic physiology in Heliotropium provide an important contribution to the study of C3-C4 intermediacy in eudicots. One of the central questions in C3-C4 biology is whether or not these species represent an intermediate state in the evolution of the C4 pathway. In 2005, a detailed, species-level phylogeny of Flaveria was published (McKown et al. 2005) characterizing the taxonomic relationships of twelve intermediate species in the genus displaying a wide range of C4-cycle activities. The

144

completed phylogeny produced several results of interest with regards to C4 evolution. First,

C3 photosynthesis is unambiguously the ancestral condition in Flaveria, and C3-C4 and C4 species are derived. Also, while there is a great deal of physiological diversity among the intermediates in Flaveria, only two, the Type II F. ramosissima and the C4-like F. palmeri, branch between C3 and C4 species (McKown et al. 2005). While this does establish C3-C4 photosynthesis as an intermediate state in the evolution of the C4 pathway, it also demonstrates that much of the photosynthetic diversity present in Flaveria intermediates is a product of radiation among the C3-C4 species rather than an indication of a stepwise progression towards the C4 syndrome. Consequently, it is necessary to develop other systems to understand in detail the progression towards C4 photosynthesis in various eudicot lineages, and to assess the extent to which these groups have followed different paths or are perhaps constrained to follow similar paths in the evolution of the C4 pathway.

This study of Heliotropium provides a potentially valuable companion to Flaveria for the study of C4 evolution. There is a great deal of geographical and functional diversity in

Heliotropium, with several dozen C4 species and over 100 C3 species in the genus, spread across Africa, southern and central Asia, Australia and North and South America (Frohlich

1978; Akhani 2007; Sage & Frohlich, unpublished). It is apparent that a species-level phylogeny of Heliotropium section Orthostachys, which contains the C4 and C3-C4 species

(Frohlich 1978; Hilger & Diane 2003), has the potential to yield valuable information about the geographical and evolutionary origins of C4 photosynthesis in this cosmopolitan genus and the significance of C3-C4 intermediacy thereto. While there is evidence to suggest that

C3 is the ancestral condition in this genus (Diane et al. 2002), it is necessary to further assess the species-level taxonomic relationships in section Orthostachys. Also, H. procumbens may

145 represent a parallel evolution of GDC restriction to the bundle-sheath based on the absence of floral bracts in this species, which led Johnston (1928) to place it in a separate, predominantly C3 clade from the other C3-C4 species of section Orthostachys included in this study. I participated in a phylogenetic assessment of Heliotropium section Orthostachys from 2006-2007 and, while the results of this study are not yet complete, there is evidence to support the assertion that C3 photosynthesis is the ancestral state in Heliotropium and that H. procumbens likely represents a parallel evolution of GDC localization in the genus (Frohlich et al, unpublished). Also, a potentially intermediate species, H. filiforme, branches at the base of an otherwise entirely-C4 clade, and this preliminary finding may support the hypothesis that C3-C4 photosynthesis was an intermediate state in the evolution of the C4 pathway in Heliotropium (Frohlich et al., unpublished).

Additional C3-C4 species are proposed to occur in Heliotropium (Frohlich 1978). In at least three species, H. lagoense, H. filiforme and H. cremnogenum, there is evidence of enlarged bundle-sheath cells and greater bundle-sheath chloroplast and mitochondria density in these species (Frohlich 1978; Hilger & Diane 2003); however, they clearly are not C4

13 species given the C3-like δ C signatures of their leaf tissues (Sage & Frohlich, unpublished).

This implies that they are also likely C3-C4 intermediates. Therefore, further study of photosynthetic gas exchange and patterns of enzyme localization are required in

Heliotropium to determine if other species in this genus exhibit C3-C4 intermediacy and whether there is evidence of C4-cycle activity in these species. If these are confirmed to be

C3-C4 intermediates, then Heliotropium would contain the second largest number of C3-C4 intermediate species of any plant taxon after Flaveria.

146

II.) Water- and nitrogen-use efficiencies of C3-C4 Flaveria species

Chapter Three of this thesis addressed the water-use and nitrogen-use efficiencies of

C3-C4 species relative to C3 and C4 species. I found that PWUE and PNUE do not increase incrementally with C4-cycle activity, but rather they shift from fully-C3 to fully-C4 levels only when CO2 assimilation is accomplished almost entirely through the C4 pathway in the

C4-like intermediates. There was no increase in PWUE or PNUE in C3-C4 species that assimilated up to 50% of their CO2 through the C4 pathway. The reasons for a wholesale, rather than incremental, increase in these values are likely the degree of integration between the C3 and C4 cycles in the Type II intermediates and the subsequent effects on carboxylation efficiency and stomatal regulation.

In C4 species, PWUE and PNUE are enhanced by the C4 CO2-concentrating mechanism. More CO2 is fixed for a given increase in Ci than in C3 leaves in the Rubisco- limited portion of the A/Ci response (Holaday et al. 1984; Rajendrudu et al. 1986; Ku et al.

1991; Vogan et al. 2007), and this increase in carboxylation efficiency results in reduced stomatal conductance and leaf Rubisco content, resulting in roughly a doubling of PWUE and PNUE (Brown 1978; Bolton & Brown 1978; Schmitt & Edwards 1981 Sage & Pearcy

1987; Li 1993; Sage 2001; Vogan et al. 2007). In Type II C3-C4 species, the presence of a partial C4 cycle reduces photorespiratory CO2 loss from the levels of Type I intermediates based on the lower Γ of these species (Ku et al. 1983; Holaday et al. 1984; Ku et al. 1991), yet this has not resulted in any significant increase in PWUE and PNUE. This result is

13 consistent with measurements of δ C, which also show no change from C3 levels until a nearly-complete C4 cycle is in place in the C4-like intermediates (Monson & Rawsthorne

2000). The degree of integration between the C3 and C4 cycles in the C4-like intermediates

147 may be a key factor in enhancement of carboxylation efficiency and, consequently, PWUE and PNUE. In C4-like intermediates, nearly all Rubisco and decarboxylating enzymes are localized to bundle-sheath cells, while PEPC is almost completely restricted to mesophyll cells (Cheng et al. 1988; Moore et al. 1989). This is not the case in Type II intermediates such as F. ramosissima (Moore et al. 1988), and the relatively more homogenous distribution of these enzymes through F. ramosissima leaves may produce some amount of futile C4- cycling in the mesophyll in which CO2 is fixed by PEPC and subsequently decarboxylated in the mesophyll, reducing the efficiency of carbon transfer to the C3 cycle from fully-C4 levels

(Monson et al. 1986; Edwards & Ku 1987; Cheng et al. 1988; Monson & Rawsthorne 2000).

This would consequently reduce carboxylation efficiency relative to species with fully- compartmentalized C3 and C4 cycles, and prevent PWUE and PNUE from approaching C4 levels in the Type II intermediates.

In addition, the mechanism of stomatal regulation of Ci may differ as well between

Type II and C4-like intermediates. Stomata are believed to respond to the intercellular, rather than ambient, concentration of carbon dioxide (Wong et al. 1979; Mott 1988), and the intercellular CO2 concentration is thought to be relayed to stomatal guard cells via an unknown mechanism (Mott 2009). In C4 species, Ci and the stomatal resistance to CO2 diffusion is altered from C3 levels, even when C3 leaves are measured at 2% oxygen (Chapter

Three), indicating that the signalling mechanism is altered in C4 relative to C3 species. In

Type II intermediates, Ci and stomatal resistance to gas diffusion are maintained at C3 levels, indicating that stomatal behavior is not altered in Type II intermediates to maintain lower Ci as is the case in C4 species. This may also contribute to the C3 levels of PWUE in the

148 intermediates as stomatal resistance to gas diffusion, and possibly the signalling mechanism as well, are not apparently different from C3 species.

The similar carboxylation efficiencies of C3, Type I and Type II species (Ku et al.

1983; 1991) indicate that there is not likely to be a change in leaf Rubisco levels, and, consequently PNUE, from C3 levels. In accordance with this hypothesis, leaf Rubisco levels and photosynthetic Rubisco-use efficiency are also unaltered from C3 levels in the Type I and

II intermediates in Chapter Three. As in the case of PWUE, it appears that fully- compartmentalized C3 and C4 cycles, and C4-like levels of PEPC activity are required before leaf Rubisco levels drop and PNUE increases. These results generally demonstrate that at current ambient CO2 concentrations, improved carbon balance through reduction of photorespiratory CO2 loss is likely the primary benefit of C3-C4 intermediacy, and PEPC may function primarily in a CO2-scavenging role in the Type II intermediates rather than being directed towards a C4-like carbon-concentrating mechanism.

One of the main avenues for future research is a characterization of the mechanism by which stomatal guard cells respond to changes in CO2 concentration. It has generally been thought that mesophyll CO2 concentration (Ci), rather than ambient CO2 concentration (Ca), primarily dictates stomatal responses to CO2 (Wong et al. 1979, 1985; Mott 1988; Mott et al.

2008); however guard cells in epidermal peels are capable to some extent of directly sensing

CO2, albeit with much lower sensitivity than in intact leaves (Assmann 1999; Messinger et al. 2006; Mott 2009). It is possible to uncouple stomatal responses to CO2 from mesophyll photosynthetic capacity using antisense constructs that reduce expression of key photosynthetic enzymes such as Rubisco (von Caemmerer et al. 2004). In these experiments, stomatal conductance is not reduced as it is when mesophyll photosynthetic capacity is

149

altered via changes in nitrogen nutrition (Wong et al. 1979; Chapter Three). Rather, Ci is dramatically higher than wild-type when photosynthetic capacity is reduced in anti-Rubisco tobacco (von Caemmerer et al. 2004); however, stomata are still capable of responding to changes in ambient CO2 concentration similar to wild-type, leading to the suggestion that stomata do not respond to Ci at all but rather to Ca or CO2 concentration in the stomatal pore.

Future research should be aimed at resolving this conflict and determining the influence of mesophyll signalling on stomatal conductance, and to assess what molecules may be involved in this response, such as ATP (Farquhar & Wong 1984) or zeaxanthin (Zhu et al.

1998).

Since stomata appear to be regulated to maintain different levels of Ci in C3 and C4 plants (Wong et al. 1979, 1985; Chapter Three), there is reason to believe that the Ci- signalling pathway may differ between C3 and C4 species. Evaluation of differences in Ci- signalling would provide information as to how stomatal behavior has evolved to mediate the trade-off between CO2 and H2O diffusion in species where the mechanism of CO2 assimilation has been greatly altered (such as C4 and C3-C4 species). Because C4 species maintain higher levels of PEPC than C3 species (Ku et al. 1991), and because PEPC has been implicated in the regulation of stomatal conductance (Lawson 2009), it is possible that increased levels of this enzyme will result in C4 species having greater stomatal sensitivity to

Ci. For example, potato plants overexpressing PEPC have more rapid stomatal responses to

CO2 than wild-type, whereas underexpressors have slower responses to CO2 (Gehlen et al.

1996). Also, antisense lines of Amaranthus edulis with reduced amounts of PEPC exhibit slower stomatal responses to CO2 and have lower maximum stomatal conductances than wild-type plants (Cousins et al. 2007). It is then possible that the increased amounts of

150

PEPC present in C4 leaves are responsible for alterations in the way Ci is sensed in these species. Also, carbonic anhydrase, which also shows elevated levels of expression in C4 leaves, has been implicated as well in stomatal responses to CO2 (Hu et al. 2010) Further research is needed to determine the relative importance of this enzyme in the stomatal function of different photosynthetic types.

III.) The effects of low CO2 on carbon balance and photosynthetic acclimation in C3, C4 and C3-C4 species from multiple evolutionary lineages

The growth and fitness of C3 plants is diminished under low atmospheric CO2 concentrations (Polley et al. 1992; Dippery et al. 1995; Campbell et al. 2005; Ward et al.

2008), and this is thought to have been an important factor selecting for the C4 carbon- concentrating mechanism (Ehleringer et al. 1991). The enhanced ability of C3-C4 species to recapture photorespired CO2 relative to C3 species may be an important factor influencing their success in low CO2 environments (Ehleringer et al. 1997). Any discernable benefits of the C3-C4 pathway to overall plant carbon balance and resource-use efficiency at low CO2 may have been crucial in advancing the initial stages of the C4 pathway (Ehleringer et al.

1997).

In this study, I found that the ability of C3 plants to acclimate to low CO2 conditions is largely constrained, both in terms of nitrogen allocation to limiting vs. non-limiting processes such as Rubisco capacity and electron transport capacity, respectively, and in terms of stomatal acclimation to facilitate greater CO2 diffusion into the leaf through higher rates of stomatal conductance. Without any evidence of phenotypic plasticity in these traits, it is clear that C3 leaves would generally have been unable to mitigate the combined stresses of

151

high air temperature and low atmospheric CO2, and selective pressure on ways to make carbon assimilation more efficient may have been great as a consequence (Cowling 2001;

Sage & Coleman 2001). C3-C4 species are able to refix proportionally more photorespired

CO2 (Monson & Rawsthorne 2000), and, as a result, exhibited roughly 50% greater rates of

-1 net CO2 assimilation at 180 µmol mol and 37°C. Increased net CO2 assimilation rates of this magnitude have translated into similar increases in aboveground biomass in previous low

CO2 experiments (reviewed in Sage & Coleman 2001), which, in turn could result in greater reproductive output and plant fitness (Solberg 1981; Arntz et al. 1998, 2000a,b). In addition, the greater rates of net CO2 assimilation in the C3-C4 species are accomplished at similar levels of stomatal conductance and leaf Rubsico content, and, consequently PWUE and

PNUE are generally elevated above C3 levels in the C3-C4 species at low CO2 and high temperature. This also would likely improve plant fitness, especially in arid or nutrient- limited environments (Dudley 1996; Ehleringer et al. 1997; Arntz et al. 2000b), as C3-C4 plants would be able to assimilate relatively more carbon than C3 species while experiencing similar levels of transpirational water loss and investing similar amounts of leaf N in

Rubisco.

Clearly, one potential future experiment would be to measure plant growth and seed set in low CO2 and high temperature conditions to determine if there is an advantage in fitness for the C3-C4 species. A factorial experiment incorporating multiple CO2 and temperature treatments would be useful in quantifying in detail the particular impacts of these factors on plant fitness in different photosynthetic types and their relative importance in selecting for C3-C4 traits. Also, measures of plant growth should include both above- and belowground biomass in such an experiment. C3 plants have been shown to allocation

152

proportionally more biomass to aboveground components under low CO2 (Dippery et al.

1995), but this may constrain the plant’s ability to assimilate sufficient nutrients from the soil, and this acclimation response may differ between photosynthetic types. It would also be useful to measure long-term growth and fitness of different photosynthetic types under high

CO2 conditions as well, given that atmospheric CO2 levels are projected to rise to over 750

-1 µmol mol by the end of the century (Meehl et al. 2007). Such an increase would bring C3 photosynthesis much closer to CO2-saturation than it is at present (von Caemmerer et al.

1997), and, in principle, could nullify the competitive advantages of the C4 carbon- concentrating mechanism (Collatz et al. 1998). However, other factors of global change such as higher temperatures, increased aridity, intensified fire cycles and land-use change may offset this (Wand et al. 1999; Sage & Kubien 2003; Leakey et al. 2004), and further research is needed to assess the impacts of these interacting factors. C3-C4 species may experience a similar elimination of their advantage in carbon assimilation capacity over C3 species, and both C4 and C3-C4 photosynthetic types may become relatively less common in high CO2 environments of the future.

153

References

Akhani H (2007) Diversity, biogeography and photosynthetic pathways of Argusia and Heliotropium (Boraginaceae) in Southwest Asia with an analysis of phytogeographical units. Botanical Journal of the Linnean Society 155: 401-425.

Anderson LJ, Maherali H, Johnson HB, Polley HW and Jackson RB (2001) Gas exchange and photosynthetic acclimation over subambient to elevated CO2 in a C3-C4 grassland. Global Change Biology 7: 693-707.

Andrews TJ and Lorimer GH (1987) Rubisco: structure, mechanisms and prospects for improvement. In: The Biochemistry of Plants, Vol. 10, eds. MD Hatch and NK Boardman. pp. 131-218.

Apel P (1994) Water use efficiency in Flaveria and Moricandia species. Biologia Plantarum 36(2): 243-246.

Apel P and Maass I (1981) Photosynthesis in species of Flaveria: CO2 compensation 13 concentration, O2 influence on photosynthetic gas exchange and δ C values in species of Flaveria (Asteraceae). Biochemie und Physiologie der Pflanzen 176, 396-399.

Apel P, Horstmann C and Pfeffer M (1997) The Moricandia syndrome in species of the Brassicaceae: evolutionary aspects. Photosynthetica 33, 205-215.

Archibold OW (1995) Ecology of World Vegetation. London, UK: Chapman and Hall.

Arntz AM and Delph LF (2001) Pattern and process: evidence for the evolution of photosynthetic traits in natural populations. Oecologia 127: 455-467.

Arntz AM, DeLucia EH and Jordan N (1998) Contribution of photosynthetic rate to growth and reproduction in Amaranthus hybridus. Oecologia 117: 323-330.

Arntz AM, DeLucia EH and Jordan N (2000a) Fitness effects of a photosynthetic mutation across contrasting environments. Journal of Evolutionary Biology 13: 792-803.

Arntz AM, DeLucia EH and Jordan N (2000b) From fluorescence to fitness: variation in photosynthetic rate affects fecundity and survivorship. Ecology 81(9): 2567-2576.

Assmann SM (1999) The cellular basis of guard cell sensing of rising CO2. Plant, Cell and Environment 22: 629-637.

Bai J, Xu DH, Kang HM, Chen K and Wang G (2008) Photoprotective function of photorespiration in Reaumuria soongorica during different levels of drought stress in natural irradiance. Photosynthetica 46(2): 232-237.

153 154

Bauwe H (1984) Photosynthetic enzyme activities and immunofluorescence studies on the localization of ribulose-1,5-bisphosphate carboxylase/oxygenase in leaves of C3, C4 and C3- C4 intermediate species of Flaveria (Asteraceae). Biochemie und Physiologie der Pflanzen 179: 253-268.

Berendse F, Berg B and Bosatta E (1987) The effect of lignin and nitrogen on the decomposition of litter in nutrient-poor ecosystems: a theoretical approach. Canadian Journal of Botany 65: 1116-1120.

Berner RA (1994) GEOCARB II: a revised model of atmospheric CO2 over Phanerozoic time. American Journal of Science 301: 182-204.

Berry J and Bjorkman O (1980) Photosynthetic response and adaptation to temperature in higher plants. Annual Review of Plant Physiology and Plant Molecular Biology 31: 491-543.

Besnard G, Muasya AM, Russier F, Roalson EH, Salamin N and Christin PA (2009) Phylogenomics of C4 photosynthesis in sedges (Cyperaceae): multiple appearances and genetic convergence. Molecular Biology and Evolution 26: 1909-1919.

Bjorkman O (1981) Responses to different quantum flux densities. In: Physiological Plant Ecology, 12A, eds. OL Lange, PS Nobel, CB Osmond, H Ziegler. Springer-Verlag, Berlin.

Bjorkman O, Pearcy RW, Mooney H and Harrison AT (1972) Photosynthetic adaptation to high temperatures: field study in Death Valley, California. Science 175: 786-789.

Bolton JK and Brown RH (1978) Effects of nitrogen nutrition on photosynthesis and associated characteristics in C3, C4 and intermediate grass species. Plant Physiology 61: 38- 41.

Bolton JK and Brown RH (1980) Photosynthesis of grass species differing in carbon dioxide fixation pathways V. Response of Panicum maximum, Panicum milioides, and tall fescue to nitrogen nutrition. Plant Physiology 66: 97-100.

Brown RH (1978) Difference in nitrogen use efficiency in C3 and C4 plants and its implications in adaptation and evolution. Crop Science 18: 93-98.

Brown RH (1999) Agronomic implications of C4 photosynthesis. In: C4 Plant Biology, eds. RF Sage and RK Monson, pp. 473-508. Academic Press, San Diego.

Brown RH and Simmons RE (1979) Photosynthesis of grass species differing in CO2 fixation pathways. I. Water-use efficiency. Crop Science 19, 375-379.

Brown RH and Hattersley PW (1989) Leaf anatomy of C3-C4 species as related to evolution of C4 photosynthesis. Plant Physiology 91: 1543-1550.

155

Campbell CD, Sage RF, Kocacinar F and Way DA (2005) Estimation of the whole-plant CO2 compensation point of tobacco (Nicotiana tabacum L.). Global Change Biology 11: 1956- 1967.

Campbell WJ, Allen LH and Bowes G (1988) Effects of CO2 concentration on Rubisco activity, amount, and photosynthesis in soybean leaves. Plant Physiology 88: 1310-1316.

Cen YP and Sage RF (2005) The regulation of Rubisco activity in response to variation in temperature and atmospheric CO2 partial pressure in sweet potato. Plant Physiology 139: 979-990.

Cerling TE (1999) Paleorecords of C4 plants and ecosystems. In: C4 Plant Biology, eds. RF Sage and RK Monson. Academic Press, San Diego, CA. pp. 445-469.

Cerling TE, Wang Y and Quade J (1993) Expansion of C4 ecosystems as an indicator of global ecological change in the Late Miocene. Nature 361: 344-345.

Cerling TE, Harris JM, MacFadden BJ, Leakey MG, Quade J, Eisenmann V and Ehleringer JR (1997) Global vegetation change through the Miocene/Pliocene boundary. Nature 389: 153-158.

Cerling TE, Harris JM and Leakey MG (1999) Browsing and grazing in elephants: the isotope record of modern and fossil proboscideans. Oecologia 120: 364-374.

14 Chastain C and Chollet R (1989) Interspecific variation in assimilation of CO2 into C4 acids by leaves of C3, C4 and C3-C4 intermediate Flaveria species near the CO2 compensation concentration. Planta 179: 81-88.

Cheng SH, Moore BD, Edwards GE and Ku MSB (1988) Photosynthesis in Flaveria brownii, a C4-like species. Plant Physiology 87: 867-873.

Christin PA and Besnard G (2009) Two independent C4 origins in Aristidoideae (Poaceae) revealed by the recruitment of distinct phosphoenolpyruvate carboxylase genes. American Journal of Botany 96: 2234-2239.

Christin PA, Besnard G, Samaritani E, Duvall MR, Hodkinson TR, Savolainen V and Salamin N (2008a) Oligocene CO2 decline promoted C4 photosynthesis in grasses. Current Biology 18: 37-43.

Christin PA, Salamin N, Muasya AM, Roalson EH, Russier F and Besnard G (2008b) Evolutionary switch and genetic convergence on rbcL following the evolution of C4 photosynthesis. Molecular Biology and Evolution 25: 2361-2368.

Christin PA, Petitpierre B, Salamin N, Buchi L and Besnard G (2008c) Evolution of C4 phosphoenolpyruvate carboxykinase in grasses, from genotype to phenotype. Molecular Biology and Evolution 26: 357-365.

156

Collatz GJ, Berry JA and Clark JS (1998) Effects of climate and atmospheric CO2 partial pressure on the global distribution of C4 grasses: present, past and future. Oecologia 114: 441-454.

Correll DS and Johnston MC (1970) Manual of the Vascular Plants of Texas. Texas Research Foundation, Renner, TX.

Cousins AB, Baroli I, Badger MR, Ivakov A, Lea PJ, Leegood RC and von Caemmerer S (2007) The role of phosphoenolpyruvate carboxylase during C4 photosynthetic isotope exchange and stomatal conductance. Plant Physiology 145: 1006-1017.

Cowling SA (2001) Plant carbon balance, evolutionary innovation and extinction in land plants. Global Change Biology 7(3): 231-239.

Cowling SA and Sage RF (1998) Interactive effects of low atmospheric CO2 and elevated temperature on growth, photosynthesis and respiration in Phaseolus vulgaris. Plant, Cell and Environment 21: 427-435.

Dai ZY, Ku MSB and Edwards GE (1996) Oxygen sensitivity of photosynthesis and photorespiration in different photosynthetic types in the genus Flaveria. Planta 198: 563- 571.

Dawe D (2000) The contribution of rice research to poverty alleviation. In: Redesigning Rice Photosynthesis to Increase Yield, eds. JE Sheehy, PL Mitchell and B Hardy, pp. 3-12. Elsevier Science B.V., Amsterdam, The Netherlands.

Dengler NG and Nelson T (1999) Leaf structure and development in C4 plants. In: C4 Plant Biology, eds. RF Sage and RK Monson. Academic Press, San Diego, CA. pp. 133-172.

Diane N, Jacob C and Hilger HH (2003) Leaf anatomy and foliar trichomes in Heliotropiceae and their systematic relevance. Flora 198: 468-485.

Dippery JK, Tissue DT, Thomas RB and Strain BR (1995) Effects of low and elevated CO2 on C3 and C4 annuals I. Growth and biomass allocation. Oecologia 101: 13-20.

Douce R and Heldt HW (2000) Photorespiration. In: Photosynthesis: Physiology and Metabolism, eds. RC Leegood, TD Sharkey and S von Caemmerer. Kluwer Academic Publishers, The Netherlands. pp. 115-136.

Drincovich MF, Casati P and Andreo CS (2001) NADP-malic enzyme from plants: a ubiquitous enzyme involved in different metabolic pathways. FEBS Letters 490: 1-6.

Dudley SA (1996) Differing selection in plant physiological traits in response to environmental water availability: a test of adaptive hypotheses. Evolution 50(1): 92-102.

157

Edwards GE and Ku MSB (1987) Biochemistry of C3-C4 intermediates. In: The Biochemistry of Plants, Vol. 10, eds. PK Stumpf and EE Conn. Academic Press, London, UK.

Edwards GE, Franceschi VR and Voznesenskaya EV (2004) Single-cell C4 photosynthesis versus the dual-cell (Kranz) paradigm. Annual Review of Plant Biology 55: 173-196.

Ehleringer JR, Sage RF, Flanagan LB and Pearcy RW (1991) Climate change and the evolution of C4 photosynthesis. Trends in Ecology and Evolution 6(3): 95-99.

Ehleringer JE and Monson RK (1993) Evolutionary and ecological aspects of photosynthetic pathway variation. Annual Review of Ecology and Systematics 24: 411-439.

Ehleringer JR, Cerling TE and Helliker BR (1997) C4 photosynthesis, atmospheric CO2 and climate. Oecologia 112: 285-299.

Evans JR (1983) Photosynthesis and nitrogen partitioning in leaves of Triticum aestivum and related species. PhD thesis, Australian National University, Canberra.

Evans JR (1989) Photosynthesis and nitrogen relationships in leaves of C3 plants. Oecologia 78: 9-19.

Farineau J, Lelandais M and Morotgaudry JF (1984) Operation of the glycolate pathway in isolated bundle sheath strands of maize and Panicum maximum. Physiologia Plantarum 60: 208-214.

Farquhar GD and Sharkey TD (1982) Stomatal conductance and photosynthesis. Annual Review of Plant Physiology and Plant Molecular Biology 33: 317-345.

Farquhar GD and von Caemmerer S (1982) Modeling of photosynthetic response to environmental conditions. In: Physiological Plant Ecology I. Encyclopedia of Plant Physiology New Series, eds. OL Lange, PS Nobel, CB Osmond and H Ziegler, Vol. 12B, pp. 549-587. Springer-Verlag, Berlin, Germany.

Farquhar GD and Wong SC (1984) An empirical model of stomatal conductance. Australian Journal of Plant Physiology 11: 191-209.

Farquhar GD, Dubbe DR and Raschke K (1978) Gain of the feedback loop involving carbon dioxide and stomata. Plant Physiology 62: 406-412.

Farquhar GD, Ehleringer JR and Hubick KT (1989) Carbon isotope discrimination and photosynthesis. Annual Review of Plant Physiology and Plant Molecular Biology 40: 503- 537.

Farris MA and Lechowicz MJ (1990) Functional interactions among traits that determine reproductive success in a native annual plant. Ecology 71(2): 548-557.

158

Field C and Mooney HA (1986) The photosynthesis-nitrogen relationship in wild plants. In: On the Economy of Plant Form and Function, ed. TJ Givnish. Cambridge University Press, Cambridge, UK.

Fox DL and Koch PL (2003) Tertiary history of C4 biomass in the Great Plains, USA. Geology 31: 809-812.

Fox DL and Koch PL (2004) Carbon and oxygen isotopic variability in Neogene paleosol carbonates: constraints on the evolution of the C4 grasslands of the Great Plains, USA. Palaeogeography, Palaeoclimatology, Palaeoecology 207: 305-329.

Frohlich MW (1978) Systematics of Heliotropium section Orthostachys in Mexico. Ph.D. thesis, Harvard University, Cambridge, MA.

Furbank RT, Hatch MD and Jenkins CLD (2000) C4 photosynthesis: mechanism and regulation. In: Photosynthesis: Physiology and Metabolism, eds. RC Leegood, TD Sharkey and S von Caemmerer. Kluwer Academic Publishers, The Netherlands.

Furbank RT, von Caemmerer S, Sheehy J and Edwards GE (2009) C4 rice: a challenge for plant phenomics. Functional Plant Biology 36: 845-856.

Garibaldi LA, Semmartin M and Chaneton EJ (2007) Grazing-induced changes in plant composition affect litter quality and nutrient cycling in flooding Pampa grasslands. Oecologia 151: 650-662.

Gehlen J, Panstruga R, Smets H, Merkelbach S, Kleines M, Porsch P, Fladung M, Becker I, Rademacher T, Hausler RE and Hirsch HJ (1996) Effects of altered phosphoenolpyruvate carboxylase activites on transgenic C3 plant Solanum tuberosum. Plant Molecular Biology 32: 831-848.

Harvey PH and Pagel M (1991) The Comparative Method in Evolutionary Biology. Oxford University Press, Oxford.

Hattersley PW, Wong SC, Perry S and Roksandic Z (1986) Comparative ultrastructure and gas-exchange characteristics of the C3-C4 intermediate Neurachne minor S.T. Blake (Poaceae). Plant, Cell and Environment 9, 217-223.

Hayes JM (1994) Global methanothophy at the Archean-Proterozoic transition. In: Early Life on Earth, ed., S Bengston. Columbia University Press, New York. pp. 220-236.

Heckathorn SA, McNaughton SJ and Coleman JS (1999) C4 plants and herbivory. In: C4 Plant Biology, eds. RF Sage and RK Monson. Academic Press, San Diego, CA.

Hibberd JM and Quick WP (2002) Characteristics of C4 photosynthesis in stems and petioles of C3 flowering plants. Nature 415: 451-454.

159

Hickman JC (1993) The Jepson Manual: Higher Plants of California. University of California Press, Berkeley, CA.

Hikosaka K and Shigeno A (2009) The role of Rubisco and cell walls in interspecific variation in photosynthetic capacity. Oecologia 160: 443-451.

Hilger HH and Diane N (2003) A systematic analysis of Heliotropiaceae (Boraginales) based on trnL and ITS1 sequence data. Botanische Jahrbucher 125, 19-51.

Holaday AS, Lee KW and Chollet R (1984) C3-C4 intermediate species in the genus Flaveria: leaf anatomy, ultrastructure, and the effect of O2 on the CO2 compensation concentration. Planta 160, 25-32.

Hopley PJ, Weedon GP, Marshall JD, Herries AIR, Latham AG and Kuykendall KL (2007) High- and low-latitude orbital forcing of early hominid habitats in South Africa. Earth and Planetary Science Letters 256: 419-432.

Hsiao TC and Jackson RB (2000) Interactive effects of water stress and elevated CO2 on growth, photosynthesis, and water-use efficiency. In: Carbon Dioxide and Environmental Stress, eds. Y Luo and HA Mooney, pp. 3-31. Academic Press, San Diego, CA.

Hu H, Boisson-Dernier A, Israelsson-Nordstrom M, Bohmer M, Xue S, Ries A, Godoski J, Kuhn JM and Schroder JI (2010) Carbonic anhydrases are upstream regulators of CO2- controlled stomatal movements in guard cells. Nature Cell Biology 12: 87-93.

Hunt S, Smith AM and Woolhouse HW (1987) Evidence for a light-dependent system for reassimilation of photorespiratory CO2 which does not include a C4 cycle, in the C3-C4 intermediate species Moricandia arvensis. Planta 171: 227-234.

Huxman TE and Monson RK (2003) Stomatal responses of C3, C3-C4 and C4 Flaveria species to light and intercellular CO2 concentration: implications for the evolution of stomatal behaviour. Plant, Cell and Environment 26, 313-322.

Hylton CM, Rawsthorne S, Smith AM, Jones DA and Woolhouse HW (1988) Glycine decarboxylase is confined to the bundle-sheath cells of leaves of C3-C4 intermediate species. Planta 175: 452-459.

Janacek SH, Trenkamp S, Palmer B, Brown NJ, Parsley K, Stanley S, Astley HM, Rolfe SA, Quick WP, Fernie AR and Hibberd JM (2009) Photosynthesis in cells around veins of the C3 plant Arabidopsis thaliana is important for both the shikimate pathway and leaf senescence as well as contributing to plant fitness. Plant Journal 59: 329-343.

Johnson HB, Polley HW and Mayeux HS (1993) Increasing CO2 and plant-plant interactions: effects on natural vegetation. Vegetatio 104/105: 157-170.

160

Johnston I.M. (1924) Studies in the Boraginaceae I. Contributions from the Gray Herbarium 70, 1-61.

Johnston IM (1928) Studies in the Boraginaceae VII. 1. The South American species of Heliotropium. Contributions from the Gray Herbarium of Harvard University 81: 3-83.

Jordan DB and Ogren WL (1984) The CO2/O2 specificity of ribulose 1,5-bisphosphate carboxylase oxygenase: dependence on ribulose bisphosphate concentration, pH and temperature. Planta 161: 308-313.

Kadereit G, Borsch T, Weising K and Freitag H (2003) Phylogeny of Amaranthaceae and Chenopodiaceae and the evolution of C4 photosynthesis. International Journal of Plant Sciences 164: 959-986.

Kapustina LA (2001) Biodiversity, ecology and microelement composition of Kyzylkum Desert shrubs (Uzbekistan). USDA Forest Service Proceedings RMRS-P-21.

Keeley JE and Rundel PW (2003) Evolution of CAM and C4 carbon-concentrating mechanisms. International Journal of Plant Sciences 164: S55-S77.

Keeley JE and Rundel PW (2005) Fire and the Miocene expansion of C4 grasslands. Ecology Letters 8(7): 683-690.

Kellogg EA (1999) Phylogenetic aspects of the evolution of C4 photosynthesis. In: C4 Plant Biology, eds. RF Sage and RK Monson. Academic Press, San Diego, CA. pp. 411-444.

Kemp PR, Waldecker DG, Owensby CE, Reynolds JF and Virginia RA (1994) Effects of elevated CO2 and nitrogen fertilization pretreatments on decomposition of tallgrass prairie leaf litter. Plant and Soil 165: 115-127.

Kennedy RA and Laetsch WM (1974) Plant species intermediate for C3, C4 photosynthesis. Science 184: 1087-1089.

Knapp AK and Medina E (1999) Success of C4 photosynthesis in the field: lessons from communities dominated by C4 plants. In: C4 Plant Biology, eds. RF Sage and RK Monson. Academic Press, San Diego, CA.

Kocacinar F, McKown AD, Sage TL and Sage RF (2008) Photosynthetic pathway influences xylem structure and function in Flaveria (Asteraceae). Plant, Cell and Environment 31: 1363-1376.

Ku MSB and Edwards GE (1977) Oxygen inhibition of photosynthesis I. Temperature dependence and relation to O2/CO2 solubility ratio. Plant Physiology 59: 986-990.

Ku MSB, Monson RK, Littlejohn RO, Nakamota H, Fisher DB and Edwards GE (1983) Photosynthetic characteristics of C3-C4 intermediate Flaveria species I. Leaf anatomy,

161

photosynthetic responses to CO2 and O2 and activities of key enzymes in the C3 and C4 pathways. Plant Physiology 71: 944-948.

Ku MSB, Wu J, Dai Z, Scott RA, Chu C and Edwards GE (1991) Photosynthetic and photorespiratory characteristics of Flaveria species. Plant Physiology 96: 518-528.

Kubien DS, Whitney SM, Moore PV and Jesson LK (2008) The biochemistry of Rubisco in Flaveria. Journal of Experimental Botany 59: 1767-1777.

Larcher W (2003) Gas exchange in plants. In: Physiological Plant Ecology, ed. W Larcher. Springer-Verlag, Berlin.

Lawson T (2009) Guard cell photosynthesis and stomatal function. New Phytologist 181: 13- 34.

Leakey ADB, Bernacchi CJ, Dohleman FG, Ort DR and Long SP (2004) Will photosynthesis of maize (Zea mays) in the US Corn Belt increase in future CO2 rich atmospheres? An analysis of diurnal courses of CO2 uptake under free-air concentration enrichment (FACE). Global Change Biology 10: 951-962.

Leegood RC (2002) C4 photosynthesis: principles of CO2 concentration and prospects for its introduction into C3 plants. Journal of Experimental Botany 53: 581-590.

Leegood RC, Sharkey TD and von Caemmerer S (2000) Introduction. In: Photosynthesis: Physiology and Metabolism, eds. RC Leegood, TD Sharkey and S von Caemmerer. Kluwer Academic Publishers, The Netherlands.

Li MR (1993) Leaf photosynthetic nitrogen-use efficiency of C3 and C4 Cyperus species. Photosynthetica 29: 117-130.

Long SP (1999) Environmental responses. In: C4 Plant Biology, eds. RF Sage and RK Monson. Academic Press, San Diego, CA.

Lorimer GH and Andrews TJ (1981) The C2 chemo- and photorespiratory carbon oxidation cycle. In: The Biochemistry of Plants, Vol. 8, eds. MD Hatch and NK Boardman. Academic Press, New York. pp. 329-374.

Lowe DR (1994) Early environments: constraints and opportunities for early evolution. In: Early Life on Earth, ed. S Bengston. Columbia University Press, New York. pp. 24-35.

Lundell CL (1966) Flora of Texas. Dallas University Press, Southern Methodist University, Dallas, TX, USA.

Maherali H, Reid CD, Polley HW, Johnson HB and Jackson RB (2002) Stomatal acclimation over a subambient to elevated CO2 gradient in a C3/C4 grassland. Plant, Cell and Environment 25: 557-566.

162

McKown AD and Dengler NG (2007) Key innovations in the evolution of Kranz anatomy and C4 vein pattern in Flaveria (Asteraceae). American Journal of Botany 94: 382-399.

McKown AD and Dengler NG (2009) Shifts in leaf vein density through accelerated vein formation in C4 Flaveria (Asteraceae). Annals of Botany 104: 1085-1098.

McKown AD, Moncalvo JM and Dengler NG (2005) Phylogeny of Flaveria (Asteraceae) and inference of C4 photosynthesis evolution. American Journal of Botany 92(11): 1911- 1928.

Meehl GA, Stocker TF, Collins WD, Friedlingstein P, Gaye AT, Gregory JM, Kitoh A, Knutti R, Murphy JM, Noda A, Raper SCB, Watterson IG, Weaver AJ and Zhao ZC (2007) Global Climate Projections. In: Climate Change 2007: The Physical Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, eds. S Solomon, D Qin, M Manning, Z Chen, M Marquis, KB Averyt, M Tignor and HR Miller. Cambridge University Press, Cambridge, UK.

Messinger SM, Buckley TN and Mott KA (2006) Evidence for involvement of photosynthetic processes in the stomatal response to CO2. Plant Physiology 140: 771-778.

Monson RK (1989a) On the evolutionary pathways resulting in C4 photosynthesis and Crassulacean acid metabolism. Advances in Ecological Research 19: 57-110.

Monson RK (1989b) The relative contributions of reduced photorespiration, and improved water- and nitrogen-use efficiencies, to the advantages of C3-C4 intermediate photosynthesis in Flaveria. Oecologia 80: 215-221.

Monson RK (1996) The use of phylogenetic perspective in comparative plant physiology and developmental biology. Annals of the Missouri Botanical Garden 83: 3-16.

Monson RK (1999) The origins of C4 genes and evolutionary pattern in the C4 metabolic phenotype. In: C4 Plant Biology, eds. RF Sage and RK Monson. Academic Press, San Diego, CA.

Monson RK (2003) Gene duplication, neofunctionalization and the evolution of C4 photosynthesis. International Journal of Plant Sciences 164: S43-S54.

Monson RK and Moore BD (1989) On the significance of C3-C4 intermediate photosynthesis to the evolution of C4 photosynthesis. Plant, Cell and Environment 12: 689-699.

Monson RK and Jaeger CH (1991) Photosynthetic characteristics of C3-C4 intermediate Flaveria floridana (Asteraceae) in natural habitats: evidence of advantages to C3-C4 photosynthesis at high leaf temperatures.

163

Monson RK and Rawsthorne S (2000) CO2 assimilation in C3-C4 intermediate plants. In: Photosynthesis: Physiology and Metabolism, eds. RC Leegood, TD Sharkey and S von Caemmerer. Kluwer Academic Publishers, The Netherlands.

Monson RK, Moore BD, Ku MSB and Edwards GE (1986) Co-function of C3- and C4- photosynthetic pathways in C3, C4 and C3-C4 intermediate Flaveria species. Planta 168: 493- 502.

Monson RK, Ku MSB and Edwards GE (1987) Photosynthesis in Flaveria brownii Powell, a C4-like C3-C4 intermediate. Plant Physiology 85: 1063-1067.

Monson RK, Teeri JA, Ku MSB, Gurevitch J, Mets LJ and Dudley S (1988) Carbon-isotope discrimination by leaves of Flaveria species exhibiting different amounts of C3- and C4-cycle co-function. Planta 174: 145-151.

Moore BD, Franceschi VR, Cheng S, Wu J and Ku MSB (1987) Photosynthetic characteristics of the C3-C4 intermediate Parthenium hysterophorus. Plant Physiology 85, 984-989.

Moore BD, Ku MSB and Edwards GE (1987) C4 photosynthesis and light-dependent accumulation of inorganic carbon in leaves of C3-C4 and C4 Flaveria species. Australian Journal of Plant Physiology 14: 657-668.

Moore BD, Monson RK, Ku MSB and Edwards GE (1988) Activities of principal photosynthetic and photorespiratory enzymes in leaf mesophyll and bundle sheath protoplasts from the C3-C4 intermediate Flaveria ramosissima. Plant Cell Physiology 29: 999-1006.

Moore BD, Ku MSB and Edwards GE (1989) Expression of C4-like photosynthesis in several species of Flaveria. Plant, Cell and Environment 12: 541-549.

Morgan CL, Turner SR and Rawsthorne S (1993) Coordination and cell-specific distribution of the four subunits of glycine decarboxylase and serine hydroxymethyltransferase in leaves of C3-C4 intermediate species from different genera. Planta 190: 468-473.

Mott KA (1988) Do stomata respond to CO2 concentrations other than intercellular? Plant Physiology 86: 200-203.

Mott KA (2009) Opinion: stomatal responses to light and CO2 depend on the mesophyll. Plant, Cell and Environment 32: 1479-1486.

Mott KA, Sibbernsen ED and Shope JC (2008) The role of the mesophyll in stomatal responses to light and CO2. Plant, Cell and Environment 31: 1299-1306.

164

Muhaidat R (2007) Diversification of C4 photosynthesis in the eudicots: anatomical, biochemical and physiological perspectives. PhD thesis, University of Toronto, Toronto, ON, Canada.

Muhaidat RM, Sage RF and Dengler NG (2006) Identification of new C3-C4 intermediates in the genus Heliotropium (Boraginaceae). Abstract, Conference of American Society of Plant Biologists, Abs # P14007.

Muhaidat R, Sage RF and Dengler NG (2007) Diversity of Kranz anatomy and biochemistry in C4 eudicots. American Journal of Botany 94: 362-381.

Murphy KL, Burke IC, Vinton MA, Lauenroth WK, Aguiar MR, Wedin DA, Virginia RA and Lowe PN (2002) Regional analysis of litter quality in the central grassland region of North America. Journal of Vegetation Science 13: 395-402.

Nisbet EG, Grassineau NV, Howe CJ, Abell PI, Regelous M and Nisbet RER (2007) The age of Rubisco: the evolution of oxygenic photosynthesis. Geobiology 5: 311-335.

O’Leary MH (1981) Carbon isotope fractionation in plants. Phytochemistry 20:553-567.

Ogren WL (1984) Photorespiration: pathways, regulation, and modification. Annual Review of Plant Physiology and Plant Molecular Biology 35: 415-442.

Ohnishi J and Kanai R (1983) Differentiation of photorespiratory activity between mesophyll and bundle sheath cells of C4 plants I. Glycine oxidation by mitochondria. Plant and Cell Physiology 24: 1411-1420.

Olson JM and Blankenship RE (2004) Thinking about the evolution of photosynthesis. Photosynthesis Research 80: 373-386.

Osborne CP (2008) Atmosphere, ecology and evolution: what drove the Miocene expansion of C4 grasslands. Journal of Ecology 96: 35-45.

Osborne CP and Beerling DJ (2006) Nature’s green revolution: the remarkable evolutionary rise of C4 plants. Philosophical Transactions of the Royal Society B – Biological Sciences 361: 173-194.

Osmond CB (1981) Photo-respiration and photoinhibition: some implications for the energetics of photosynthesis. Biochimica et Biophysica Acta 639: 77-98.

Ott L (1977) Models relating a response to a set of independent variables. In: An Introduction to Statistical Methods and Data Analysis (ed. L. Ott), pp. 458-519. Duxbury Press, Belmont, CA.

Pagani M, Freeman KH and Arthur MA (1999) Late Miocene atmospheric CO2 concentrations and the expansion of C4 grasses. Science 285: 876-878.

165

Passey BH, Cerling TE, Perkins ME, Voorhies MR, Harris JM and Tucker ST (2002) Environmental change in the Great Plains: an isotopic record from fossil horses. Journal of Geology 110: 123-140.

Pearson PN and Palmer MR (2000) Atmospheric carbon dioxide concentrations over the past 60 million years. Nature 406: 695-699.

Phelan PL and Stinner BR (1992) Microbial mediation of plant-herbivore ecology. In: Herbivores – Their Interactions with Secondary Plant Metabolites, eds. GA Rosenthal and ME Berenbaum. Academic Press, San Diego, CA.

Polley HW, Johnson HB and Mayeux HS (1992) Growth and gas exchange of oats (Avena sativa) and wild mustard (Brassica kaber) at subambient CO2 concentrations. International Journal of Plant Sciences 153: 453-461.

Polley HW, Johnson HB, Mayeux HS and Malone SR (1993a) Physiology and growth of wheat across a subambient carbon dioxide gradient. Annals of Botany 71: 347-356.

Polley HW, Johnson HB, Marino BD and Mayeux HS (1993b) Increase in C3 plant water-use efficiency and biomass over Glacial to present CO2 concentrations. Nature 361: 61-64.

Polley HW, Johnson HB and Derner J (2002) Soil- and plant-water dynamics in a C3/C4 grassland exposed to a subambient to superambient CO2 gradient. Global Change Biology 8: 1118-1129.

Porra RJ (2006) Spectrometric assays for plant, algal and bacterial chlorophylls. In: Chlorophylls and Bacteriophylls: Biochemistry, Biophysics, Functions and Applications, eds. B Grimm, RJ Porra, W Rudiger and H Scheer. Springer Publishing, The Netherlands.

Powell AM (1978) Systematics of Flaveria (Flaveriinae – Asteraceae). Annals of the Missouri Botanical Garden 65: 590-636.

Pulquério MJF and Nichols RA (2007) Dates from the molecular clock: how wrong can we be? Trends in Ecology and Evolution 22: 180-184.

Quade J and Cerling TE (1995) Stable isotopes in paleosols and the expansion of C4 grasses in the late Miocene of Northern Pakistan. Palaeogeography, Palaeoclimatology, Palaeoecology 115: 91-116.

Rajendrudu G, Prasad JSR and Das VSR (1986) C3-C4 intermediate species in Alternanthera (Amaranthaceae): leaf anatomy, CO2 compensation point, net CO2 exchange and activities of photosynthetic enzymes. Plant Physiology 80: 409-414.

Rawsthorne S (1992) C3-C4 intermediate photosynthesis: linking physiology to gene expression. Plant Journal 2: 267-274.

166

Rawsthorne S, Hylton CM, Smith AM and Woolhouse HW (1988) Photorespiratory metabolism and immunogold localization of photorespiratory enzymes in leaves of C3 and C3-C4 intermediate species of Moricandia. Planta 173: 298-308.

Reed JE and Chollet R (1985) Immunofluorescent localization of phosphoenolpyruvate carboxylase and ribulose 1,5-bisphosphate carboxylase/oxygenase proteins in leaves of C3, C4 and C3-C4 Flaveria species. Planta 165: 439-445.

Ripley BS, Abraham TI and Osborne CP (2008) Consequences of C4 photosynthesis for the partitioning of growth: a test using C3 and C4 subspecies of Alloteropsis semialata under nitrogen-limitation. Journal of Experimental Botany 59: 1705-1714.

Ross DJ, Tate KR, Newton PCD and Clark H (2002) Decomposability of C3 and C4 grass litter sampled under different concentrations of atmospheric carbon dioxide at a natural CO2 spring. Plant and Soil 240: 275-286.

Roth-Nebelsick A, Uhl D, Mosbrugger V and Kerp H (2001) Evolution and function of leaf venation architecture: a review. Annals of Botany 87: 553-566.

Royer DL (2006) CO2-forced climate thresholds during the Phanerozoic. Geochemica et Cosmochemica Acta 70: 5665-5675.

Royer DL, Wing SL, Beerling DJ, Jolley DW, Koch PL, Hickey LJ and Berner RA (2001) Paleobotanical evidence for near present-day levels of atmospheric CO2 during part of the Tertiary. Science 292: 2310-2313.

Rumpho ME, Ku MSB, Cheng SH and Edwards GE (1984) Photosynthetic characteristics of C3-C4 intermediate Flaveria species III. Reduction of photorespiration by a limited C4 pathway of photosynthesis in Flaveria ramosissima. Plant Physiology 75: 993-996.

Sage RF (1999) Why C4 photosynthesis? In: C4 Plant Biology, eds. RF Sage and RK Monson. Academic Press, San Diego, CA.

Sage RF (2001) C4 plants. In: Encyclopedia of Biodiversity, Vol. 1, ed. SA Levin. Academic Press, San Diego, CA.

Sage RF (2002) Variation in the kcat of Rubisco in C3 and C4 plants and some implications for photosynthetic performance at high and low temperature. Journal of Experimental Botany 53: 609-620.

Sage RF (2004) The evolution of C4 photosynthesis. New Phytologist 161: 341-370.

Sage RF and Pearcy RW (1987a) The nitrogen use efficiency of C3 and C4 plants I. Leaf nitrogen, growth, and biomass partitioning in Chenopodium album (L.) and Amaranthus retroflexus (L.). Plant Physiology 84: 954-958.

167

Sage RF and Pearcy RW (1987b) The nitrogen use efficiency of C3 and C4 plants II. Leaf nitrogen effects on the gas exchange characteristics of Chenopodium album (L.) and Amaranthus retroflexus (L.). Plant Physiology 84: 959-963.

Sage RF and Reid CD (1992) Photosynthetic acclimation to sub-ambient CO2 (20 Pa) in the C3 annual Phaseolus vulgaris L. Photosynthetica 27: 605-617.

Sage RF and Seemann JR (1993) Regulation of ribulose-1,5-bisphosphate carboxylase/oxygenase activity in response to reduced light intensity in C4 plants. Plant Physiology 102: 21-28.

Sage RF and Reid CD (1995) Photosynthetic response mechanisms to environmental change in C3 plants. In: Plant-Environment Interactions, ed. RE Wilkinson. Marcel Dekker, Inc., New York.

Sage RF and Pearcy RW (2000) The physiological ecology of C4 photosynthesis. In: Photosynthesis: Physiology and Metabolism, eds. RC Leegood, TD Sharkey and S von Caemmerer. Kluwer Academic Publishers, The Netherlands.

Sage RF and Coleman JR (2001) Effects of low atmospheric CO2 on plants: more than a thing of the past. Trends in Plant Science 6: 18-24.

Sage RF and Kubien DS (2003) Quo vadis C4? An ecophysiological perspective on global change and the future of C4 plants. Photosynthesis Research 77: 209-225.

Sage RF, Pearcy RW and Seemann JR (1987) The nitrogen use efficiency of C3 and C4 plants III. Leaf nitrogen effects on the activity of carboxylating enzymes in Chenopodium album (L.) and Amaranthus retroflexus (L.). Plant Physiology 85: 355-359.

Sage RF, Reid CD, Moore BD and Seemann JR (1993) Long-term kinetics of the light- dependent regulation of ribulose-1,5-bisphosphate carboxylase oxygenase activity in plants with and without 2-carboxyarabitinol 1-phosphate. Planta 191: 222-230.

Sage RF, Li M and Monson RK (1999) The taxonomic distribution of C4 photosynthesis. In: C4 Plant Biology (eds. R. F. Sage & R. K. Monson), pp. 551-584. Academic Press, San Diego.

Sage RF, Wedin DA and Li MR (1999) The biogeography of C4 photosynthesis. In: C4 Plant Biology, eds. RF Sage and RK Monson. Academic Press, San Diego, CA.

Samson R, Mani S, Boddey R, Sokhansanj S, Queseda D, Urguiaga S, Reis V and Lem CH (2005) The potential of C4 perennial grasses for developing global BIOHEAT industry. Critical Reviews in Plant Sciences 24, 461-495.

168

Sanchez-del Pino I (2009) Phylogeny and subgeneric classification of the genus Alternanthera (Amaranthaceae, Gomphrenoideae) and its multiple colonizations of Galapagos Islands. PhD thesis, Freien Universitat Berlin, Berlin, Germany.

Sayre RT and Kennedy RA (1977) Ecotypic differences in C3 and C4 photosynthetic activity in Mollugo verticillata, a C3-C4 intermediate. Planta 134, 257-262.

Schmitt MR and Edwards GE (1981) Photosynthetic capacity and nitrogen use efficiency of maize, wheat and rice: a comparison between C3 and C4 photosynthesis. Journal of Experimental Botany 32: 459-466.

Schulze ED and Hall AE (1982) Stomatal responses, water loss and CO2 assimilation rates of plants in contrasting environments. In: Physiological Plant Ecology, 12B, eds. OL Lange, PS Nobel, CB Osmond, H Ziegler. Springer-Verlag, Berlin.

Schulze ED, Ellis R, Schulze W and Trimborn P (1996) Diversity, metabolic types and δ13C carbon isotope ratios in the grass flora of Namibia in relation to growth form, precipitation and habitat conditions. Oecologia 106: 352-369.

Schuster WS and Monson RK (1990) An examination of the advantages of C3-C4 intermediate photosynthesis in warm environments. Plant, Cell and Environment 13: 903- 912.

Sharkey TD (1988) Estimating the rate of photorespiration in leaves. Physiologia Plantarum 73: 147-152.

Sheehy J, Mitchell PL, Hardy B (2007) Charting New Pathways to C4 Rice. World Scientific Publishing Company, Hackensack, NJ.

Solbrig OT (1981) Studies on the population ecology of the genus Viola II. The effect of plant size on fitness in Viola sororia. Evolution 35(6): 1080-1093.

Somerville CR and Ogren WL (1979) A phosphoglycolate phosphatase-deficient mutant of Arabidopsis. Nature 280: 833-835.

Somerville CR and Ogren WL (1982) Mutants of the cruciferous plant Arabidopsis thaliana lacking glycine decarboxylase activity. Biochemical Journal 202: 373-380.

Stowe LG and Teeri JA (1978) Geographic distribution of C4 species of dicotyledonae in relation to climate. American Naturalist 112: 609-623.

Stromberg AE (2005) Decoupled taxonomic radiation and ecological expansion of open- habitat grasses in the Cenozoic of North America. PNAS 102: 11980-11984.

169

Sudderth EA, Muhaidat R, McKown AD, Kocacinar F and Sage RF (2007) Leaf anatomy, gas exchange and photosynthetic enzyme activity in Flaveria kochiana. Functional Plant Biology 34, 118-129.

Svensson P, Blasing OE and Westhoff P (2007) Evolution of C4 phosphoenolpyruvate carboxylase. Archives of Biochemistry and Biophysics 414: 180-188.

Tabita FR, Hanson TE, Li HY, Satagopan S, Singh J and Chan S (2007) Function, structure and evolution of the RubisCO-like proteins and their RubisCO homologs. Microbiology and Molecular Biology Reviews 71: 576-597.

Tabita FR, Satagopan S, Hanson TE, Kreel NE and Scott SS (2008) Distinct form I, II, III and IV Rubisco proteins from the three kingdoms of life provide clues about Rubisco evolution and structure/function relationships. Journal of Experimental Botany 59: 1515- 1524.

Teeri JA and Stowe LG (1976) Climatic patterns and the distribution of C4 grasses in North America. Oecologia 23: 1-12.

Thomas RB and Strain BR (1991) Root restriction as a factor in photosynthetic acclimation of cotton seedlings grown in elevated carbon dioxide. Plant Physiology 96: 627-634.

Tidwell WD and Nambudiri EMV (1989) Tomlinsonia thomassonii Gen. et Sp. Nov., a permineralized grass from the Upper Miocene Ricardo Formation, California. Review of Palaeobotany and Palynology 60: 165-177.

Tipple BJ and Pagani M (2007) The early origins of terrestrial C4 photosynthesis. Annual Review of Earth and Planetary Sciences 35: 435-461.

Tissue DT, Griffin KL, Thomas RB and Strain BR (1995) Effects of low and elevated CO2 on C3 and C4 annuals II. Photosynthesis and biochemistry. Oecologia 101: 21-28.

Tripati AK, Roberts CD and Eagle RA (2009) Coupling of CO2 and ice sheet stability over major climate transitions of the last 20 million years. Science 326: 1394-1397.

Ueno O (2001) Ultrastructural localization of photosynthetic and photorespiratory enzymes in epidermal, mesophyll, bundle sheath and vascular bundle cells of the C4 dicot Amaranthus viridis 52: 1003-1013.

Van Dyne GM , Brockington NR, Szocs Z, Duek J and Ribic CA (1980) Large herbivore subsystem. In: Grasslands, Systems Analysis and Man, eds. AI Breymeyer and GM Van Dyne. Cambridge University Press, Cambridge, UK.

Vincentini A, Barber JC, Aliscioni SA, Giussani LM and Kellogg EA (2008) The age of grasses and clusters of origins of C4 photosynthesis. Global Change Biology 14: 2963-2977.

170

Vogan PJ, Frohlich MW and Sage RF (2007) The functional significance of C3-C4 intermediate traits in Heliotropium: gas exchange perspectives. Plant, Cell and Environment 30: 1337-1345. von Caemmerer S. (1989) A model of photosynthetic CO2 assimilation and carbon-isotope discrimination in leaves of certain C3-C4 intermediates. Planta 178, 463-474. von Caemmerer S (2000) Biochemical Models of Leaf Photosynthesis. CSIRO Publishing, Collingwood, Australia. von Caemmerer S and Hubick KT (1989) Short-term carbon-isotope discrimination in C3-C4 intermediate species. Planta 178, 475-481. von Caemmerer S and Quick WP (2000) Rubisco: physiology in vivo. In: Photosynthesis: Physiology and Metabolism, eds. RC Leegood, TD Sharkey and S von Caemmerer. Kluwer Academic Publishers, The Netherlands. von Caemmerer S and Furbank RT (2003) The C4 pathway: an efficient CO2 pump. Photosynthesis Research 77: 191-207. von Caemmerer S, Millgate A, Farquhar GD and Furbank RT (1997) Reduction of ribulose- 1,5-bisphosphate carboxylase/oxygenase by antisense RNA in the C4 plant Flaveria bidentis leads to reduced assimilation rates and increased carbon isotope discrimination. Plant Physiology 113: 469-477. von Caemmerer S, Lawson T, Oxborough K, Baker NR, Andrews TJ and Raines CA (2004) Stomatal conductance does not correlate with photosynthetic capacity in transgenic tobacco with reduced amounts of Rubisco. Journal of Experimental Botany 55: 1157-1166.

Voznesenskaya EV, Artyusheva EG, Franceschi VR, Pyankov VI, Kiirats O, Ku MSB and Edwards GE (2001) Salsola arbusculiformis, a C3-C4 intermediate in Salsoleae (Chenopodiaceae). Annals of Botany 88, 337-348.

Voznesenskaya EV, Koteyeva NK, Chuong SDX, Ivanova AN, Barroca J, Craven LA and Edwards GE (2007) Physiological, anatomical and biochemical characterization of photosynthetic types in genus Cleome (Cleomaeceae). Functional Plant Biology 34: 247-267.

Wan CSM and Sage RF (2001) Climate and the distribution of C4 grasses along the Atlantic and Pacific coasts of North America. Canadian Journal of Botany 79: 474-486.

Wand SJE, Midgley GF, Jones MH and Curtis PS (1999) Responses of wild C4 and C3 grass (Poaceae) species to elevated atmospheric CO2 concentration: a meta-analytic test of current theories and perceptions. Global Change Biology 5: 723-741.

Ward JK, Tissue DT, Thomas RB and Strain BR (1999) Comparative responses of model C3 and C4 plants to drought in low and elevated CO2. Global Change Biology 5: 857-867.

171

Ward JK, Antonovics J, Thomas RB and Strain BR (2000) Is atmospheric CO2 a selective agent on model C3 annuals? Oecologia 123: 330-341.

Ward JK, Harris JM, Cerling TE, Wiedenhoeft A, Lott MJ, Dearing MD, Coltrain JB and Ehleringer JR (2005) Carbon starvation in glacial trees recovered from the La Brea tar pits, Southern California. PNAS 102: 690-694.

Ward JK, Myers DA and Thomas RB (2008) Physiological and growth responses of C3 and C4 plants to reduced temperature when grown at low CO2 of the last ice age. Journal of Integrative Plant Biology 50(11): 1388-1395.

Webster GL, Brown WV and Smith BN (1975) Systematics of photosynthetic carbon fixation pathways in Euphorbia. Taxon 24: 27-33.

Wedin DA and Tilman D (1990) Species effects on nitrogen cycling: a test with perennial grasses. Oecologia 84: 433-441.

Wedin DA and Tilman D (1993) Competition among grasses along a nitrogen gradient: initial conditions and mechanisms of competition. Ecological Monographs 63: 199-229.

Wedin DA and Tilman D (1996) Influence of nitrogen loading and species composition on the carbon balance of grasslands. Science 274: 1720-1723.

Wingler A, Lea PJ, Quick WP and Leegood RC (2000) Photorespiration: metabolic pathways and their role in stress protection. Philosophical Transactions of the Royal Society B – Biological Sciences 355: 1517-1529.

Wong SC, Cowan IR and Farquhar GD (1979) Stomatal conductance correlates with photosynthetic capacity. Nature 282: 424-426.

Wong SC, Cowan IR and Farquhar GD (1985) Leaf conductance in relation to rate of CO2 assimilation I. Influence of nitrogen nutrition, phosphorus nutrition, photon flux density, and ambient partial pressure of CO2 during ontogeny. Plant Physiology 78: 821-825.

Yoshimura Y, Kubota F and Ueno O (2004) Structural and biochemical bases of photorespiration in C4 plants: quantification of organelles and glycine decarboxylase. Planta 220: 307-317.

Zachos J, Pagani M, Sloan L, Thomas E and Billups K (2001) Trends, rhythms and abberations in global climate 65 Ma to present. Science 292: 686-693.

Zhu J, Talbott LD, Jin X and Zeiger E (1998) The stomatal response to CO2 is linked to changes in guard cell zeaxanthin. Plant, Cell and Environment 21: 813-820.