Carbon Inputs to Ecosystems 5

Home , Atriplex hymenelytra, Stomatal conductance, Tradescantia

Photosynthesis by plants provides the carbon address carbon inputs to ecosystems through and energy that drive most biological pro- photosynthesis in this chapter and the carbon cesses in ecosystems. This chapter describes losses from plants and ecosystems in Chaps. 6 the controls over carbon input to ecosystems. and 7, respectively. The balance of these processes governs the patterns of carbon accumulation and loss in ecosystems and the carbon distribution Introduction between the land, atmosphere, and ocean.

The energy fixed by photosynthesis directly supports plant growth and produces organic A Focal Issue matter that is consumed by animals and soil microbes. The carbon derived from photosynthe- Carbon and water exchange through pores sis makes up about half of the organic matter on (stomata) in the leaf surface governs the effi- Earth; hydrogen and oxygen account for most of ciency with which increasingly scarce water the rest. Human activities have radically modified resources support food production for a grow- the rate at which carbon enters the terrestrial bio- ing human population. Open stomata (Fig. 5.1) sphere by changing most of the controls over this maximize carbon gain and productivity when process. We have increased the quantity of atmo- water is abundant, but at the cost of substantial spheric CO2 by 35% to which terrestrial plants are water loss. Partial closure of stomata under dry exposed. At regional and global scales, we have conditions reduces carbon gain but increases the altered the availability of water and nutrients, the efficiency with which water supports plant growth. major soil resources that determine the capacity What constrains the capacity of the biosphere to of plants to use atmospheric CO2. Finally, through gain carbon? Where and in what seasons does changes in land cover and the introduction and most photosynthesis occur? How do plants regu- extinction of species, we have changed the late the balance between carbon gain and water regional distribution of the carbon-fixing potential loss? Application of current understanding of the of the terrestrial biosphere. Because of the central controls over tradeoffs between carbon gain and role that carbon plays in the climate system (see water loss could reduce the likelihood of a “train Chap. 2), the biosphere, and society, it is critical wreck” resulting from current trends in increasing that we understand the factors that regulate food demands and declining availability of fresh- its cycling through plants and ecosystems. We water to support agricultural production.

F.S. Chapin, III et al., Principles of Terrestrial Ecosystem Ecology, 123 DOI 10.1007/978-1-4419-9504-9_5, © Springer Science+Business Media, LLC 2011 124 5 Carbon Inputs to Ecosystems

Fig. 5.1 Surface of a Tradescantia virginiana leaf with opening influences both the maximum rate and the effi- open stomatal pores. Selection for plants that differ in sto- ciency with which plants use water to gain carbon. matal density and physiological regulation of stomatal Photograph courtesy of Peter Franks

Carbon is the main element that plants reduce Overview of Carbon Inputs with energy derived from the sun. Carbon and to Ecosystems energy are therefore tightly linked as they enter, move through, and leave ecosystems. Photosynthesis is the process by which most Photosynthesis uses light energy (i.e., radiation carbon and chemical energy enter ecosystems. in the visible portion of the spectrum) to reduce

The proximate controls over photosynthesis at CO2 and produce carbon-containing organic the cellular or leaf level are the availability of compounds. This organic carbon and its associ- photosynthetic reactants such as light energy ated energy are then transferred among compo- and CO2; temperature, which governs reaction nents within the ecosystem and are eventually rates; and the availability of nitrogen, which is released to the atmosphere by respiration or required to produce photosynthetic enzymes. combustion. Photosynthesis at the scale of ecosystems is The energy content of organic matter varies termed gross primary production (GPP). Like among carbon compounds, but for whole tissues, photosynthesis by individual cells or leaves, GPP it is relatively constant at about 20 kJ g−1 of ash- varies diurnally and seasonally in response to free dry mass (Golley 1961; Larcher 2003; variations in light, temperature, and nitrogen sup- Fig. 5.3). The carbon concentration of organic ply. Differences among ecosystems in annual matter is also variable but averages about 45% of GPP, however, are determined primarily by the dry weight in herbaceous tissues and 50% in wood quantity of photosynthetic tissue and the duration (Gower et al. 1999; Sterner and Elser 2002). Both of its activity (Fig. 5.2). These, in turn, depend on the carbon and energy contents of organic matter the availability of soil resources (water and nutri- are greatest in materials such as seeds and animal ents), climate, and time since disturbance. In this fat that have high lipid content and are lowest in chapter, we explore the mechanisms behind these tissues with high concentrations of minerals or causal relationships. organic acids. Because of the relative constancy Biochemistry of Photosynthesis 125

LONG-TERM SHORT-TERM CONTROLS CONTROLS

STATE Interactive Direct FACTORS controls controls

Plant BIOTA functional Leaf area types

TIME

Soil Nitrogen resources

Season GPP PARENT length MATERIAL

Temperature CLIMATE Light

CO2

Fig. 5.2 The major factors governing temporal and Thickness of the arrows indicates the strength of the direct spatial variation in gross primary production (GPP) in and indirect effects. The factors that account for most of ecosystems. These controls range from proximate con- the variation among ecosystems in GPP are leaf area and trols, which determine the diurnal and seasonal variations length of the photosynthetic season, which are ultimately in GPP, to the interactive controls and state factors, which determined by the interacting effects of soil resources, are the ultimate causes of ecosystem differences in GPP. climate, vegetation, and disturbance regime

of the carbon and energy contents of organic Needle-leaved matter, carbon, energy, and biomass have been 28 tree Broad-leaved used interchangeably as currencies of the carbon tree and energy dynamics of ecosystems. The pre- ) 26 Herb

− 1 ferred units differ among subfields of ecology, depending on the processes that are of greatest 24 interest or are measured most directly. Production studies, for example, typically focus on biomass, 22 trophic studies on energy, and gas exchange studies on carbon. Energy content (kJ g 20

18 Biochemistry of Photosynthesis

Leaf Stem Seed Root/ The biochemistry of photosynthesis governs rhizome the environmental controls over carbon inputs Fig. 5.3 Energy content of major tissues in conifer trees, to ecosystems. Photosynthesis involves two major broad-leaved trees, and broad-leaved herbs. Compounds groups of reactions: The light-harvesting reac- that contribute to a high energy content include lipids tions (or light-dependent reactions) transform light (seeds), terpenes and resins (conifers), proteins (leaves), and lignin (woody tissues). Values are expressed per gram energy into temporary forms of chemical energy of ash-free dry mass. Data from Larcher (2003) (ATP and NADPH; Lambers et al. 2008). The 126 5 Carbon Inputs to Ecosystems

Light Light-harvesting Starch reactions e chl Sugar export NADP

+ 2 3C sugars H O H + O2 2 Carbon-fixation reactions oid oid ylak

ylak NADPH CO2 Th Th RuBP O2 ATP (5C sugar)

H+ Photo- respiration 2 2C compounds ADP Thylakoid 2C Stroma export

Fig. 5.4 A chloroplast, showing the location of the major provide the energy to synthesize ribulose-bisphosphate photosynthetic reactions. The light-harvesting reactions (RuBP), which reacts either with CO2 to produce sugars and occur in the thylakoid membranes; chlorophyll (chl) absorbs starch (carbon-fixation reactions of photosynthesis) or with visible light and funnels the energy to reaction centers, O2 to produce two-carbon intermediates (photorespiration) + where water inside the thylakoid is split to H and O2, and and ultimately CO2. Through either carbon fixation or pho- resulting electrons are passed down an electron-transport torespiration, ADP and NADP are regenerated to become chain in the thylakoid membrane, ultimately to NADP, pro- reactants in the production of additional ATP and NADPH. ducing NADPH. During this process, protons move across The net effect of photosynthesis is to convert light energy the thylakoid membrane to the stroma, and the proton (H+) into chemical energy (sugars and starches) that is available gradient drives the synthesis of ATP. ATP and NADPH to support plant growth and maintenance carbon-fixation reactions (or light-independent potential efficiency of photosynthesis in convert- reactions, sometimes called the dark reaction) ing solar radiation into chemical energy. use the products of the light-harvesting reactions The carbon-fixation reactions of photosynthe- to convert CO2 into sugars, a more permanent sis use the chemical energy (ATP and NADPH) form of chemical energy that can be stored, from the light-harvesting reactions to reduce CO2 transported, or metabolized. Both groups of to sugars. The rate-limiting step in the carbon- reactions occur simultaneously in the light in fixation reactions is the reaction of a five-carbon chloroplasts, which are organelles inside pho- sugar (ribulose-bisphosphate [RuBP]) with CO2 to tosynthetic cells (Fig. 5.4). In the light-harvest- form two three-carbon organic acids (phospho- ing reactions, chlorophyll (a light-absorbing glycerate), which are then reduced using ATP and pigment) captures energy from visible light. NADPH from the light reactions to form three- Absorbed radiation is converted to chemical carbon sugars (glyceraldehyde 3-phosphate). The energy (NADPH and ATP), and oxygen is pro- initial attachment of CO2 to a carbon skeleton is duced as a waste product. Visible radiation catalyzed by the enzyme ribulose-bisphosphate accounts for 40% of incoming solar radiation carboxylase-oxygenase (Rubisco). The rate of this (see Chap. 2), which places an upper limit on the reaction is generally limited by the products of the Biochemistry of Photosynthesis 127 light-harvesting reaction and by the concentration one three-carbon acid (phosphoglycerate), which of CO2 in the chloroplast. A surprisingly high can be recycled back to RuBP. If the plant were to concentration of Rubisco is required for carbon acquire this phosphoglycerate solely through fixation. Rubisco accounts for about 25% of the assimilation of three new CO2 molecules, the cost nitrogen in photosynthetic cells, and other pho- would be 9 ATP and 6 NADPH. Photorespiration tosynthetic enzymes make up an additional 25%. may also act as a safety valve by providing a sup- The remaining enzymatic steps in the carbon- ply of reactants (ADP and NADP) to the light fixation reactions use ATP and NADPH from the reaction under conditions in which an inade- light-harvesting reactions to convert some mole- quate supply of CO2 limits the rate at which these cules of the three-carbon sugar (glyceraldehyde reactants can be regenerated by carbon-fixation 3-phosphate) to RuBP, thus closing the photosyn- reactions. In the absence of photorespiration, thetic carbon reduction cycle, and convert the rest continued light harvesting produces oxygen radi- to the six-carbon sugar, glucose, that is transported cals that destroy photosynthetic pigments. out of the chloroplast (Fig. 5.4). The most notable Plants have additional lines of defense against features of the carbon-fixation reactions are: (1) excessive energy capture that are at least as impor- their large nitrogen requirement for Rubisco and tant as photorespiration. Terrestrial plants and other photosynthetic enzymes; (2) their depen- algae in shallow coral reefs, for example, have a dence on the products of the light-harvesting reac- photoprotection mechanism involving changes tions (ATP and NADPH), which in turn depend on in pigments of the xanthophyll cycle. When exci- irradiance, i.e., the light received by the photo- tation energy in the light-harvesting reactions synthetic cell; and (3) their frequent limitation by exceeds the capacity of these reactions to synthe-

CO2 delivery to the chloroplast. The basic bio- size ATP and NADPH, the xanthophyll pigment is chemistry of photosynthesis therefore dictates that converted to a form that receives this excess this process must be sensitive to light and CO2 absorbed energy from the excited chlorophyll and availability over timescales of milliseconds to dissipates it harmlessly as heat (Demming-Adams minutes and sensitive to nitrogen supply over tim- and Adams 1996). This processing of excess escales of days to weeks (Fig. 5.2; Evans 1989). energy under high light prevents photodestruc- Rubisco is both a carboxylase, which initiates tion of photosynthetic pigments under these the carbon-fixation reactions of photosynthesis, conditions. and an oxygenase, which catalyzes the reaction The photosynthetic reactions described above between RuBP and oxygen (Fig. 5.4). Early in are known collectively as C3 photosynthesis the evolution of photosynthesis on Earth, oxygen because two molecules of the three-carbon acid, concentrations were very low, and CO2 concen- phosphoglycerate are the initial products of car- trations were high, so the oxygenase activity of bon fixation. C3 photosynthesis is the fundamen- this enzyme occurred at negligible rates (Sage tal photosynthetic pathway of all photosynthetic 2004). The oxygenase initiates a series of steps organisms on Earth, although there are impor- that break down sugars to CO2. This process of tant variations on this theme that we discuss photorespiration immediately respires away later. Plant chloroplasts, for example, have many 20–40% of the carbon fixed by photosynthesis similarities to, and probably evolved from, sym- and regenerates ADP and NADP in the process. biotic bluegreen photosynthetic bacteria. Other Why do plants have such an inefficient system of carbon-fixation reactions contribute to the pho- carbon acquisition, by which they immediately tosynthesis of some terrestrial plants (C4 photo- lose a third of the carbon that they acquire through synthesis and Crassulacian Acid Metabolism photosynthesis? Photorespiration is best viewed or CAM). These reactions initially produce a as a carbon recovery process. Photorespiration four-carbon acid that is subsequently broken recycles about 75% of the carbon processed by down to release CO2 that enters the normal the oxygenase activity of Rubisco at a cost of two C3 photosynthetic pathway to produce three-

ATPs and one NADPH to produce one CO2 and carbon sugars. However, the bottom line is that 128 5 Carbon Inputs to Ecosystems

Limited by C fixation C3 photosynthesis is the fundamental mechanism by which carbon enters all ecosystems, so an ) Light understanding of its environmental controls − 1 Light Photo- s limitation saturation oxidation

provides considerable insight into the carbon − 2 dynamics of ecosystems. Net photosynthesis is the net rate of carbon gain measured at the level of individual cells or leaves. It is the balance between simultaneous Light CO2 fixation and respiration of photosynthetic compensation cells in the light (including both photorespira- point tion and mitochondrial respiration). Respiration 0 1000 2000 rate is proportional to protein content, so photo- Net photosynthesis ( µ mol m Irradiance (µmol m−2 s−1) synthetic cells and leaves with a high capacity for photosynthesis (lots of photosynthetic pro- Fig. 5.5 Relationship of net photosynthetic rate to photo- tein), also lose a lot of carbon due to their high synthetically active radiation and the processes that limit photosynthesis at different irradiances. The linear increase respiration rate. The light compensation point in photosynthesis in response to increased light (in the (irradiance at which photosynthesis just balances range of light limitation) indicates relatively constant respiration) is therefore higher in cells or leaves light-use efficiency. The light compensation point is the that have a high photosynthetic capacity. There minimum irradiance at which the leaf shows a net gain of carbon is therefore a tradeoff between the capacity of plants to photosynthesize at high light (lots of protein and high photosynthetic capacity) and their performance at low light (less protein, increases linearly with increasing light (Fig. 5.5). lower respiration rate, and positive net photosyn- The slope of this line (the quantum yield of pho- thesis at low light availability, i.e., a low light tosynthesis) is a measure of the efficiency with compensation point). which photosynthetic cells use absorbed light to Plants adjust the components of photosyn- produce sugars. Quantum yield is similar (about thesis, so the energy trapped by light-harvesting 1–4% of the incoming light energy) among all C3 reactions closely matches the energy needed plants (both aquatic and terrestrial) at low light in for the CO2-fixation reactions. As plants pro- the absence of environmental stress (Kalff 2002; duce new cells over days to weeks, protein syn- Lambers et al. 2008). At high irradiance, photo- thesis is distributed between light-harvesting vs. synthesis becomes light saturated, i.e., it no longer carbon-fixing enzymes so that capacities for light responds to changes in light supply, due to the harvesting and carbon fixation are approximately finite capacity of light-harvesting reactions to balanced under the typical light and CO2 environ- capture light. As a result, light energy is converted ment of the cell or leaf. Plants increase their less efficiently into sugar energy at high light. investment in light-harvesting capacity in low- Photosynthetic capacity (maximum photosyn- light environments and their carbon-fixing capac- thetic rate measured at light saturation) depends ity at high light. Total photosynthetic capacity on the quantity of photosynthetic enzymes in the reflects the quantity of photosynthetic enzymes, cell and is generally higher in large-celled algal which depends on nitrogen acquisition from their species and rapidly growing terrestrial species environment. Once a photosynthetic cell is pro- that characterize nutrient-rich waters and lands, duced, there is limited capacity to adjust the respectively. Photosynthesis declines at extremely proportions of light-harvesting and carbon-fixing high light, when the xanthophylls cycle photo- enzymes. protective process in the chloroplast are over- At low light, where the supply of ATP and whelmed, due to photo-oxidation of photosynthetic NADPH from the light-harvesting reactions lim- enzymes and pigments (Kalff 2002; Mann and its the rate of carbon fixation, net photosynthesis Lazier 2006; Lambers et al. 2008). Pelagic Photosynthesis 129

In the next sections, we describe how environ- clear water of the open-ocean and oligotrophic mental controls over photosynthesis operate in (low-nutrient) lakes, water accounts for most of aquatic and terrestrial ecosystems. We begin with the energy absorption, and high-energy blue light aquatic systems, where most primary producers penetrates to the greatest depth, up to 50–100 m in are single-celled organisms (phytoplankton), and clear lakes (Kalff 2002) and 200 m in the open water seldom limits photosynthesis, thus simpli- ocean (Fig. 5.6; Valiela 1995). In eutrophic (high- fying the nature of environmental controls over nutrient) lakes and rivers, chlorophyll absorbs carbon entry to the ecosystem. We then add most of the light, which may penetrate only a few the additional complexities found in terrestrial meters or less. Tannins absorb most light in tea- ecosystems. colored oligotrophic lakes in acidic low-nutrient landscapes. The depth of light penetration has two important consequences for pelagic ecosystems. First, it determines the depth of the euphotic Pelagic Photosynthesis zone, where there is enough light to support phytoplankton growth, i.e., where their photosynthe- Light Limitation sis exceeds respiration (see Chap. 6). This is often defined arbitrarily as the depth at which light is Photosynthesis in pelagic (open-water) ecosys- 1% of that available at the surface, although some tems of lakes and the ocean depends on light phytoplankton photosynthesis occurs at even availability and phytoplankton biomass. Light lower light intensities (Kalff 2002). In small, enters water at the surface of lakes and the ocean shallow lakes, which are by far the most numer- and decreases exponentially with depth: ous, the euphotic zone extends to the lake bot-

-kz tom, and much of the production occurs on the I= Ie (5.1) z o lake bottom, particularly in nutrient-poor settings where I is the irradiance (the quantity of radiant (Vadeboncoeur et al. 2002; Vander Zanden et al. energy received at a surface per unit time) at depth 2006; Vadeboncoeur et al. 2008). Second, the z (m), Io is the irradiance at the water surface; and depth of light penetration in lakes influences k is the extinction coefficient. Light reduction stratification because most of the absorbed solar through the water column results from absorp- radiation is converted to heat, which reduces tion by water, chlorophyll, dissolved organic sub- water density and promotes stratification (warmer stances, and organic or sediment particles. In the less dense water at the surface). Eutrophic lakes

0 Terrestrial forest

200 Coastal ocean water

400

600 Clearest oceanLower water limit of Depth (m) 800 phytoplankton Fig. 5.6 Light availability at different distances beneath the surface of a 1000 Lower limit forest canopy (Chazdon of fish and Fetcher 1984) and the 1200 coastal and open ocean −13 −9 −5 −1 3 10 10 10 10 10 (Valiela 1995). Modified −2 from Valiela (1995) Irradiance (W m ) 130 5 Carbon Inputs to Ecosystems

a toms with silica skeletons sink more rapidly than Z < Z Productivity Biomass eu mix other phytoplankton (Kalff 2002; Mann and Lazier 2006). Phytoplankton are like the terrestrial shade

Zeu plants that will be described later. Due to their relatively low concentration of photosynthetic Zmix enzymes, they have both a low photosynthetic capacity and a low respiration rate. They there- b Z = Z fore maintain positive net photosynthesis at the eu mix low light levels that characterize most of the water column and the depths at which cells spend most of their lives. Maximum photosynthesis in

Zeu= Zmix marine phytoplankton typically occurs at 5–25% of full sun, a few meters below the water surface (Valiela 1995; Mann and Lazier 2006). High light intensities that occur near the water surface on c Zeu > Zmix clear days reduce photosynthetic rate, but, due to turbulent mixing, phytoplankton spend relatively little time near the surface. Below the depth of

Zmix maximum photosynthesis, carbon uptake declines

Zeu with depth in parallel with the exponential decline Aphotic in light intensity. The depth of the euphotic zone is often similar to or less than the mixing depth of surface

Fig. 5.7 Influence of the relative depths of the euphotic waters. In this case, there is a relatively uniform zone (zeu) and mixed layer (zmix) on the vertical distribution depth distribution of phytoplankton biomass, of phytoplankton and biomass. Redrawn from Thornton and the depth distribution of photosynthesis et al. (1990) can be readily predicted from the light response curve of photosynthesis and the depth profile of with shallow light penetration therefore tend to light availability (Fig. 5.7b). In strongly strati- show greatest stratification and are most resistant fied or extremely clear lakes, light sometimes to wind-driven mixing. penetrates more deeply than the mixed layer. In The distribution of photosynthesis through the this case, there is an additional peak in phyto- water column depends on the depth distribution plankton biomass and photosynthesis at the base of phytoplankton and their photosynthetic of the euphotic zone driven by the greater nutri- response to light intensity (Valiela 1995; Kalff ent availability below the mixed layer (Fig. 5.7c). 2002). Mixing of the surface water typically The actual depth distribution of photosynthesis occurs more rapidly (e.g., an hour or less) than is more complex than these simple rules imply phytoplankton can produce new cells (about a because variability in mixing creates vertical and day; see Fig. 2.21), so turbulent mixing rather horizontal patchiness in the distribution of nutri- than cellular production or death determines the ents and phytoplankton. vertical distribution of phytoplankton and there- In the ocean and clear lakes at high latitudes, fore the depth distribution of photosynthetic UV-B may also contribute to low photosynthetic potential in the water column (Fig. 5.7; Thornton rates in surface waters, raising questions about et al. 1990). When winds are calm and in sheltered whether aquatic production may have been lakes, other factors that influence the vertical reduced by high-latitude increases in UV-B (the distribution of phytoplankton include rates of cell “ozone holes” caused by anthropogenic CFCs; production and mortality and the rates at which see Chap. 1). Colored dissolved organic com- algae sink or swim. Large-bodied algae and dia- pounds absorb UV-B radiation, so changes in Pelagic Photosynthesis 131 these dissolved organics will likely mediate any marine pelagic ecosystems, for example, only potential UV-B impacts on aquatic ecosystems 1% of the carbon in a given water volume is (Williamson et al. 1996; Kalff 2002). Photosynthesis involved in primary production, whereas the at the ocean or lake surface appears to be light- nitrogen in this water may cycle through primary limited mainly at high latitudes during winter due production 10–100 times a year (Thurman 1991). to low solar angles, short days, and snow-covered One reason for the apparently low responsiveness ice. At depth, light limits photosynthesis in all of pelagic photosynthesis to carbon supply is pelagic habitats. that inorganic carbon is available in substantial

concentrations in several forms, including CO2, bicarbonate, carbonate, and carbonic acid. When

CO2 Supply CO2 dissolves in water, a small part is transformed to carbonic acid, which in turn dissociates to Photosynthesis is less often carbon-limited in bicarbonate, carbonate, and H+ ions with a aquatic than in terrestrial ecosystems. In concomitant drop in pH.

+ - +-2 H2 O+ CO 2 « H 23 CO «+ H HCO3 « 2H + CO3 (5.2)

As expected from these equilibrium reactions, just as in the ocean. Groundwater entering the predominant forms of inorganic carbon are freshwater ecosystems is super-saturated with free CO2 and carbonic acid at low pH (the equa- CO2 derived from root and microbial respiration tion driven to the left), soluble bicarbonate at in terrestrial soils (Kling et al. 1991; Cole et al. about pH 8 (typical of ocean waters), and carbon- 1994). Most streams, rivers, and lakes are net ates at high pH (equation driven to the right). sources of CO2 to the atmosphere because the

Fossil-fuel emissions to the atmosphere have CO2 input from groundwater generally exceeds increased the CO2 inputs to the ocean, driving the capacity of aquatic primary producers to use

(5.2) to the right. The resulting 30% increase in the CO2. In addition, aquatic decomposition of ocean acidity (H+) tends to dissolve the carbonate both aquatic and terrestrially derived organic shells of marine invertebrates and calcareous carbon generates a large CO2 source within lakes phytoplankton (coccolithophores) with potentially and rivers (see Chap. 7; Kortelainen et al. 2006; profound impacts on the functioning of marine Cole et al. 2007). Eutrophic lakes with their ecosystems (see Chap. 14). Bicarbonate accounts high plankton biomass have a greater demand for 90% of the inorganic carbon in most marine for CO2 to support photosynthesis than do olig- waters. Despite the predominance of bicarbonate otrophic systems, but their organic accumula- in the ocean, phytoplankton in pelagic ecosystems tion and high decomposition rate in sediments use CO2 as their primary carbon source. As CO2 also contribute a large CO2 input to the water is consumed, it is replenished from bicarbonate column from depth. This creates a strong verti-

(5.2). Some marine algae in the littoral zone, such cal gradient in CO2 in stratified eutrophic lakes, as the macroalga, Ulva, also use bicarbonate. with CO2 being absorbed from the atmosphere It is still actively debated the extent to which during the day and returned at night (Carpenter marine productivity will respond directly to et al. 2001), just as in terrestrial ecosystems. increasing atmospheric CO2. Phytoplankton with Some freshwater vascular plants such as Isoetes low affinity for bicarbonate and most phyto- use CAM photosynthesis to acquire CO2 at plankton under eutrophic conditions increase night and refix it by photosynthesis during the photosynthesis and growth in response to added day (Keeley 1990). Other freshwater vascular

CO2 (Schippers et al. 2004). plants transport CO2 from the roots to the can-

Daily photosynthesis in unpolluted fresh- opy to supplement CO2 supplied from the water water ecosystems is seldom carbon-limited, column. 132 5 Carbon Inputs to Ecosystems

Nutrient Limitation Pelagic GPP

Nutrients limit phytoplankton photosynthesis Total photosynthesis of pelagic ecosystems inte- primarily through their effects on the produc- grates the effects of nutrients on phytoplankton tion of new cells. Productivity and photosynthe- biomass and the effects of light and other envi- sis are closely linked in all ecosystems through a ronmental factors on the photosynthetic activ- system of amplifying (positive) feedbacks (see ity of individual cells. GPP is the rate of Chap. 6): Photosynthesis provides the carbon photosynthesis integrated through the water col- and energy to produce new photosynthetic cells, umn, typically over time steps of days to a year which increases the quantity of photosynthesis (e.g., g C m−2 of ecosystem yr−1). Ecosystem mod- that can occur. This feedback is particularly eling and remote sensing have played a major role strong in pelagic systems, where most primary in estimating GPP in aquatic ecosystems. Turbulent production is by phytoplankton through the pro- mixing maintains a relatively homogeneous distri- duction of new photosynthetic cells. Nutrients bution of photosynthetic capacity throughout the strongly limit productivity in most unpolluted surface mixed layer (constant photosynthetic aquatic ecosystems, both freshwater and marine. capacity and light compensation point), although As nutrient availability increases, the rate of pro- the efficiency with which chlorophyll traps light duction of new cells increases but each cell adjusts relatively rapidly and is greater at depth maintains a relatively modest concentration of than at the surface (Flynn 2003; Mann and Lazier photosynthetic enzymes, which accounts for 2006). Because of the relatively homogeneous their low photosynthetic capacity and low light photosynthetic capacity through the mixed layer, compensation point. In other words, phytoplank- chlorophyll content is a useful indicator of phyto- ton respond to nutrient supply primarily by plankton biomass. In the ocean, the vertical distri- increasing photosynthetic biomass, not by bution of light absorption by chlorophyll can be increasing the photosynthetic capacity of indi- estimated from satellite-derived color images of vidual cells. This increases the amount of phyto- the ocean surface using SeaWiFS (Sea-viewing plantkon biomass distributed through the water Wide Field-of-view Sensor). SeaWiFS estimates column but enables each cell to function in the the depth profile of radiation absorbed by chloro- low-light environment in which it spends most phyll because different wavelengths of light pen- of its life (due to its low light compensation etrate to different depths. point, which is a consequence of its low photo- As discussed earlier, the shape of photosynthe- synthetic capacity). sis-depth curve depends on the intensity and depth Phytoplankton species differ somewhat in of turbulent mixing and the depth of light penetra- photosynthetic capacity. Large-celled species tion (Fig. 5.7; Thornton et al. 1990; Kalff 2002; with a high photosynthetic capacity dominate Mann and Lazier 2006). Lakes accumulate carbon eutrophic waters, whereas small-celled nano- when the total photosynthesis integrated through plankton (2–20 mm in diameter) and pico- the water column (GPP) exceeds the total respira- plankton (<2 mm in diameter) dominate tion. The compensation depth is the depth at which oligotrophic waters. As described in Chaps. 6 GPP equals phytoplankton respiration integrated and 9, large-celled species have an advantage in through the water column. If the mixing depth is producing biomass rapidly when nutrients are below the compensation depth, phytoplankton res- readily available. In contrast, small-celled spe- piration beneath this depth exceeds photosynthesis, cies, with their higher surface-to-volume ratio, and they lose carbon. In the most productive pelagic are less limited by nutrient diffusion to the cell ecosystems, such as eutrophic lakes and upwelling surface and are competitively favored in nutrient- systems, the mixing depth is considerably shal- poor waters. lower than the compensation depth. Living on the Edge: Streams and Shorelines 133

plants and mosses, benthic (bottom-dwelling) Living on the Edge: Streams algae, epiphytic algae that attach to the surface and Shorelines of vascular plants, moss and macroalgae, and planktonic algae that float in slow-moving waters. Streams and littoral (shoreline) habitats have The relative contribution of different primary pro- properties that depend on both terrestrial and ducers to photosynthesis differs among geomorphic aquatic components. On the terrestrial side, zones (erosional, transfer, and depositional) riparian vegetation benefits from a stable water within the river basin and depends on patterns of supply and what is often a relatively favorable flow rate, flood frequency, and substrate stability nutrient environment (see Chap. 13; Naiman (see Chap. 3). Small headwater streams in the et al. 2005). For this reason, salt marshes, fresh- erosional zone of a drainage basin are often water marshes, and emergent vegetation along shaded by riparian vegetation, have relatively stable lakeshores often support high rates of pho- high flows (at least in some seasons), and vari- tosynthesis and productivity (Valiela 1995). On able nutrient inputs, depending on the dynamics the aquatic side, shading by emergent vascular of adjacent terrestrial ecosystems. Attached algae plants and terrestrial vegetation largely defines (periphyton), mosses, and liverworts on rocks the light environment of headwater streams and and stable sediments generally account for most stable lake and stream banks, as described later. of the photosynthesis in headwater streams (Allan Lotic (flowing-water) ecosystems such as and Castillo 2007). As headwater streams join to streams and rivers have unique properties that form larger rivers, the greater solar input supports distinguish them from both lakes and terrestrial more photosynthesis by macrophytes along shal- systems. Primary producers of streams include low stable riverbanks and by periphyton on stable macrophytes (large plants) such as vascular riverbeds (Fig. 5.8). During periods of low flow,

Control sites 8 0 2.4 1.2 1

6 − )

) 0.8 − 1 −3 ) − 1 4

− 1 yr yr

− 2

− 2 − 2

0.4 2 m −5 (kg C C (kg 00 −7 10−2 10−0 102 104 10−2 10−0 102 104 10−2 10−0 102 104

Human-dominated sites 8 0 1.2 6 −1 Ecosystem respiration (kg C m Gross primary production (kg C m 0.8 production ecosystem Net 4 −3

0.4 2 −5

0 0 −7 10−1 101 103 105 10−2 10−0 102 104 10−2 10−0 102 104

Watershed area (km2) [log scale]

Fig. 5.8 Gross primary production, ecosystem respiration, and net ecosystem production of rivers and streams that differ in watershed area. Redrawn from Finlay 2011 134 5 Carbon Inputs to Ecosystems benthic algae such as Cladophora can form side channels, lakes, reservoirs, or pools and get extensive mats (Power 1992b). Benthic mosses swept into the river channel. Since the maximum are important in many cold water streams and doubling time of most phytoplankton is once or rivers. In slow-moving rivers polluted by waste- twice per day, there is a strong inverse relation- water or agricultural runoff, pelagic algae can ship between discharge and phytoplankton bio- dominate if the doubling rate of algae is more mass in rivers. River phytoplankton populations rapid than their rate of downstream export (Allan can be self-sustaining if the currents are slow and Castillo 2007). In general, GPP increases enough and nutrients are abundant enough to with increasing stream size, although it is quite support rapid production throughout the year. In variable, especially in large human-dominated other cases, the rivers are seasonally seeded with drainage basins (Finlay 2011). phytoplankton from river-associated lakes and The controls over photosynthesis in streams side channels. The roles of light and nutrients in and rivers vary depending on primary pro- controlling photosynthesis of river phytoplank- ducer type and environment. Benthic algae in ton are similar to those in lakes. The total photo- forested headwater streams, for example, have synthesis (GPP) in a section of river depends not relatively low rates of photosynthesis because of only on the light environment and photosynthetic low light availability, just as on the adjacent ter- properties of the plants in that ecosystem but also restrial forest floor. Removal of riparian trees and on algal transport from upstream river segments, shrubs often increases photosynthesis and pro- as discussed in Chaps. 7 and 9. duction in deforested headwater streams (Allan and Castillo 2007). In other cases, nutrients so strongly limit algal growth that algae show rela- Terrestrial Photosynthesis tively little response to added light. In general, nutrients influence benthic photosynthesis pri- Photosynthetic Structure marily through their effect on the rate of produc- of Terrestrial Ecosystems tion of new photosynthetic cells rather than on the photosynthetic properties of those cells, just The physical differences between air and as in lakes and the ocean. As discussed later, the water account for the major photosynthetic high turbulence of flowing waters reduces limita- differences between terrestrial and aquatic tion by nutrient diffusion to algal cells, so nutrient ecosystems. Aquatic algae are bathed in water limitation tends to be less pronounced in flowing that physically supports them and brings CO2 water than in pelagic ecosystems. Because of the and nutrients directly to photosynthetic cells. super-saturation of groundwater with CO2, pho- Water turbulence continuously mixes planktonic tosynthesis in the streams that receive this algae to different positions in the vertical light groundwater is seldom CO2-limited. gradient. In contrast, the leaves of terrestrial Stream macrophytes generally contribute a rel- plants are suspended from elaborate support atively small proportion of the photosynthetic car- structures and remain at fixed locations in the bon inputs to flowing-water ecosystems because canopy. These leaves and their support structures of the small proportion of the stream surface area create and respond to the vertical light gradient that they usually occupy. Mosses tend to dominate in terrestrial canopies. Thus, in contrast to phy- in shaded headwater streams, especially when toplankton, terrestrial leaves have opportunities waters are cold, and floating or emergent vascular to adjust photosynthesis to a particular light plants dominate in lowland floodplain rivers and environment. Photosynthetic cells in the leaves estuaries with slower currents, greater sediment of terrestrial plants are encased in waxy cuticles accumulation, and higher light availability. to minimize water loss, but this impermeable

The phytoplankton present in the water col- coating also slows CO2 diffusion to the sites of umn of slow-moving eutrophic rivers often origi- carbon fixation in chloroplasts. Terrestrial leaves nate from permanent populations in slow-moving thus face tradeoffs between water loss and Terrestrial Photosynthesis 135

Day Night

H O CO CO C3 plants 2 2 2

C3 Ps

Rmi

Mes Mes Sugar C4 plants CO2 CO2 H2O

C4 Ps C3 + Rmi Rmi C4 acid CO2 C Ps Mes 3 BS Mes BS

Sugar CO2 H2O CAM plants C acid C Ps 4 C44 Ps

CO2 + C3

C3 Ps C4 acid Mes Mes Sugar

Fig. 5.9 Cellular location and diurnal timing of CO2 fixa- fixation (C4 Ps) occurs in mesophyll cells and C3 fixation tion and water exchange in leaves with C3, C4, and (C3 Ps) occurs in bundle sheath (BS) cells. Mitochondrial

Crassulacean Acid Metabolism (CAM) photosynthetic respiration (Rmi) occurs at night. Exchanges with the pathways. In C3 and CAM plants, all photosynthesis atmosphere of CO2 and water vapor occur during the day occurs in mesophyll (Mes) cells. In C4 plants, C4 carbon in C3 and C4 plants and at night in CAM plants

CO2 absorption that are not an issue in aquatic the short distance from the cell surface to the ecosystems. chloroplast. C3 leaf chloroplasts contain an

In terrestrial plants, the CO2 used in photo- enzyme, carbonic anhydrase that catalyzes the synthesis diffuses along a concentration gradient conversion of bicarbonate to dissolved CO2, from the atmosphere outside the leaf to the chlo- maximizing the concentration of the form of car- roplast. CO2 first diffuses across a layer of rela- bon (CO2) that is fixed by Rubisco. The bound- tively still air close to the leaf surface (the leaf ary layer, stomata, and cellular water all influence boundary layer) and then through the stomata the overall diffusion of CO2 from the free air to (small pores in the leaf surface), the diameter of Rubisco, but stomata are the largest (and most which is regulated by the plant (Figs. 5.1, 5.9; variable) component of this resistance. The thin,

Lambers et al. 2008). Once inside the leaf, CO2 flat shape of most leaves and the abundance of diffuses through air spaces between cells, dis- air spaces inside leaves maximize the rate of CO2 solves in water on the cell surfaces, and diffuses diffusion from the bulk air to the chloroplast. 136 5 Carbon Inputs to Ecosystems

Cell walls inside the leaf are coated with a (Still et al. 2003). C4 species dominate many thin film of water that facilitates the efficient warm, high-light environments, particularly trop- transfer of CO2 from the air to the interior of ical grasslands and savannas. C4-dominated eco- cells. This water readily evaporates, and water systems account for nearly a third of the ice-free vapor diffuses out through the stomata across the terrestrial surface (see Table 6.6) and are therefore boundary layer to the atmosphere. The open sto- quantitatively important in the global carbon mata that are necessary for plants to gain carbon cycle. In C4 photosynthesis, phosphoenolpyru- are therefore also an avenue for water loss (see vate (PEP) is first carboxylated byPEP carboxy- Chap. 4). In other words, terrestrial plants face lase in mesophyll cells to produce four-carbon an inevitable tradeoff between CO2 absorption organic acids (Fig. 5.9). These organic acids are (which is necessary to drive photosynthesis) and transported to specialized bundle sheath cells, water loss (which must be replaced by absorp- where they are decarboxylated. The CO2 released tion of water from the soil). This tradeoff can be from the organic acids then enters the normal C3 as high as 400 molecules of water lost for each pathway of photosynthesis to produce sugars that molecule of CO2 absorbed. Plants regulate CO2 are exported from the leaf. There are three eco- absorption and water loss by changing the size of logically important features of the C4 photosyn- stomatal openings, which regulates stomatal thetic pathway: conductance, the flux of water vapor, or CO2 per First, C4 acids move to the bundle sheath cells, unit driving force (i.e., for a given concentration where they are decarboxylated, concentrating gradient). When plants reduce stomatal conduc- CO2 at the site where Rubisco fixes carbon. This tance to conserve water, photosynthesis declines, increases the efficiency of carboxylation by reducing the efficiency with which plants convert Rubisco because it increases the concentration light energy to carbohydrates. Plant regulation of of CO2 relative to O2, which would otherwise

CO2 delivery to the chloroplast is therefore a compete for the active site of the enzyme. compromise between maximizing photosynthe- Apparent photorespiration measured at the leaf sis and minimizing water loss and depends on level is low in C4 plants because most of the the relative supplies of CO2, light, and mineral RuBP in the bundle sheath chloroplasts reacts nutrients, as described later. We now describe with CO2 rather than with O2 and because the two photosynthetic pathways that enhance plant PEP carboxylase in the mesophyll cells scav-

performance in warm, high-light environments enges any photorespired CO2 that diffuses away

(C4 photosynthesis) and dry environments (CAM from the bundle sheath cells. photosynthesis). Second, PEP carboxylase draws down the

concentration of CO2 inside the leaf to a greater

extent than does Rubisco. This increases the CO2

C4 Photosynthesis concentration gradient between the external air

and the internal air spaces of the leaf. A C4 plant

C4 photosynthesis adds an additional set of can therefore absorb CO2 with more tightly carbon-fixation reactions that enable some closed stomata than can a C3 plant, thus reducing plants to increase net photosynthesis in warm, water loss. high-light environments by reducing photo- Third, the net cost of regenerating the carbon respiration. About 85% of vascular-plant spe- acceptor molecule (PEP) of the C4 pathway is cies fix carbon by the 3C photosynthetic pathway, two ATPs for each CO2 fixed, a 30% increase in in which Rubisco is the primary carboxylating the energy requirement of photosynthesis com- enzyme. The first biochemically stable products pared to C3 plants. of C3 photosynthesis are three-carbon organic The major advantage of the C4 photosyn- acids. About 3% of the global flora photosynthe- thetic pathway is increased carboxylation sizes by the C4 photosynthetic pathway (Sage under conditions that would otherwise favor 2004), contributing about 23% of terrestrial GPP photorespiration (Sage 2004). Due to their lack Terrestrial Photosynthesis 137 of photorespiration, which increases exponentially Crassulacean Acid Metabolism with rising temperature, C4 plants maintain higher rates of net photosynthesis at high temperatures Crassulacean acid metabolism (CAM) is a than do C3 plants; this explains the success of C4 photosynthetic pathway that enables plants plants in warm environments. C4 photosynthesis to gain carbon under extremely dry condi- initially evolved with similar frequency in mesic, tions. Succulent plant species (e.g., cactuses) in arid, and saline environments, and today’s C4 dry environments, including many epiphytes in plants appear to be no more drought tolerant than the canopies of tropical forests, gain carbon

C3 plants (Sage 2004). Nonetheless, the low sto- through CAM photosynthesis. CAM accounts matal conductance of C4 plants appears to pre- for a small proportion of terrestrial carbon gain adapt them to dry conditions, so C4 genera now because it is active only under extremely dry occur in a wider range of dry habitats than their C3 conditions. Even in these environments, some counterparts (Osborne and Freckleton 2009). The CAM plants switch to C3 photosynthesis when main disadvantage of the C4 pathway is the addi- enough water is available. tional energy cost for each carbon fixed by photo- In CAM photosynthesis, plants close their synthesis, which is best met under high-light stomata during the day, when high tissue tem- conditions (Edwards and Smith 2010). The C4 peratures and low relative humidity of the exter- pathway is therefore most advantageous in warm, nal air would otherwise cause large transpirational high-light conditions, such as tropical grasslands water loss (Fig. 5.9). At night, they open their and marshes. The C4 pathway occurs in 18 plant stomata, and CO2 enters the leaf and is fixed by families and has evolved independently at least PEP carboxylase. The resulting C4 acids are

45 times (Sage 2004). C4 species first became stored in vacuoles until the next day when they abundant in the late Miocene 6–8 million years are decarboxylated, releasing CO2 to be fixed by ago, probably triggered by a global decline in normal C3 photosynthesis. Thus, in CAM plants atmospheric CO2 concentration (Cerling 1999). there is a temporal (day-night) separation of C3

C4 grasslands expanded during glacial periods, and C4 CO2 fixation, whereas in 4C plants there is when CO2 concentrations declined, and retracted a spatial separation of C3 and C4 CO2 fixation at the end of glacial periods, when atmospheric between bundle sheath and mesophyll cells.

CO2 concentration increased, suggesting that the CAM photosynthesis is energetically expensive, evolution of C4 photosynthesis was tightly tied to like C4 photosynthesis; it therefore occurs pri- variations in atmospheric CO2 concentration. marily in dry, high-light environments such as However, there is little geographic variation in deserts, shallow rocky soils, and canopies of atmospheric CO2 concentration, so the current geo- tropical forests. CAM photosynthesis allows graphic distribution of C4 plants appears to be some plants to gain carbon under extremely dry controlled primarily by temperature and light conditions that would otherwise preclude carbon availability, rather than by CO2 concentration. fixation in ecosystems.

C4 plants have an isotopic signature that allows tracking of their past and present role in ecosystems. C4 plants incorporate a larger CO2 Limitation 13 fraction of C than do C3 plants during photosynthesis (Box 5.1) and therefore have a distinct Plants adjust the components of photosynthe- isotopic signature that characterizes any organic sis, so physical and biochemical processes co- matter that originated by this photosynthetic limit carbon fixation. Photosynthesis operates pathway, including animals and soil organic mat- most efficiently when the rate of CO2 diffusion ter. Isotopic measurements are a valuable tool in into the leaf matches the biochemical capacity of studying ecological processes in ecosystems the leaf to fix CO2. Terrestrial plants regulate the components of photosynthesis to approach this bal- where the relative abundance of C3 and C4 plants has changed over time (Ehleringer et al. 1993). ance, as seen from the response of photosynthesis 138 5 Carbon Inputs to Ecosystems

Box 5.1 Carbon Isotopes 12 13 The three isotopic forms of carbon ( C, C, Table 5.1 Representative 13C concentrations (‰) of 14 and C) differ in their number of neutrons but atmospheric CO2 and selected plant and soil materials have the same number of protons and elec- Material ∂13C (‰)a trons. The additional atomic mass causes the PeeDee limestone standard 0.0

heavier isotopes to react more slowly in some Atmospheric CO2 −8 reactions, particularly in the carboxylation of Plant material

Unstressed C3 plant −27 CO2 by Rubisco. Carboxylating enzymes pref- Water-stressed C plant −25 erentially fix the lightest of these isotopes of 3 Unstressed C plant −13 carbon (12C). C plants generally have a rela- 4 3 Water-stressed C plant −13 tively high CO concentration inside the leaf, 4 2 CAM plantb −27 to −11 due to their high stomatal conductance. Under Soil organic matter these circumstances, Rubisco discriminates Derived from unstressed C plants −27 against the heavier isotope 13C, causing 13CO 3 2 Derived from C4 or CAM plants −13 to accumulate within the airspaces of the leaf. Data from O’Leary (1988) and Ehleringer and Osmond 13 CO2 therefore diffuses out of the leaf through (1989) the stomata along a concentration gradient of a The concentrations are expressed relative to an inter- 13 12 nationally agreed-on standard (PeeDee belemnite): CO2 at the same time that CO2 is diffusing

into the leaf. In C4 and CAM plants, in con- æöR ¶=13C 1000sam - 1 trast, PEP carboxylase has such a high affinity std ç÷ èøRstd for CO2 that it reacts with most of the CO2 that enters the leaf, resulting in relatively little dis- where ∂13 C is the isotope ratio in delta units relative crimination against 13CO . Consequently, the 2 to a standard, and Rsam and Rstd are the isotope abun- 13 C concentrations of CAM and C4 plants are dance ratios of the sample and standard, respectively much higher (less negative isotopic ratios) (Ehleringer and Osmond 1989) bValues of −11 under conditions of CAM photosyn- than those of C3 plants (Table 5.1). thesis; many CAM plants switch to C3 photosynthesis This difference in isotopic composition under favorable moisture regimes, giving an isotopic

among C3, C4, and CAM plants remains in any ratio similar to that of unstressed C3 plants organic compounds derived from these plants.

This makes it possible to calculate the relative C4 dominance (or vice versa), the organic mat-

proportions of C3 and C4 plants in the diet of ter in plants differs in its isotopic composition animals by measuring the 13C content of the from that of the soil (and its previous vegeta- animal tissue; this can be done even in fossil tion). Changes in the carbon isotope composi- bones such as those of early humans. Changes tion of soil organic matter over time then in the isotopic composition of fossil bones are provides a tool to estimate the current rates of a clear indicator of changes in diet. In situa- turnover of soil organic matter that formed

tions where vegetation has changed from C3 to beneath the previous vegetation.

to the CO2 concentration inside the leaf (Fig. 5.10). leaf. The plant can increase photosynthesis only

When the internal CO2 concentration is low, pho- by opening stomatal pores. Alternatively, if CO2 tosynthesis increases approximately linearly with concentration inside the leaf is high, photosyn- increasing CO2 concentration. Under these cir- thesis shows little response to variation in CO2 cumstances, the leaf has more carbon-fixation concentration (the asymptote approached in capacity than it can use, and photosynthesis is Fig. 5.10). In this case, photosynthesis is limited limited by the rate of diffusion of CO2 into the by the rate of regeneration of RuBP (the compound Terrestrial Photosynthesis 139

) Limitation by − 1 the break point of the CO2-response curve s CO2 diffusion − 2 (Fig. 5.10; Körner et al. 1979), where CO2 supply and carbon-fixation capacity are about equally limiting to photosynthesis. Biochemical Changes in stomatal conductance by leaves limitation minimize the effects of CO2 supply on photo- CO2 synthesis. The free atmosphere is so well mixed compensation that its CO concentration varies globally by only point 2 4% – not enough to cause significant regional 0

Net photosynthesis ( µ mol m 350 700 variation in photosynthesis. In dense canopies, Internal concentration (ppmv) CO2 photosynthesis reduces CO2 concentration somewhat within the canopy, and soil respiration is a Fig. 5.10 Relationship of the net photosynthetic rate to source of CO2 at the base of the canopy. However, the CO2 concentration inside the leaf. Photosynthetic rate is limited by the rate of CO diffusion into the chloroplast the shade leaves in the lower canopy tend to be 2 light-limited and therefore relatively unrespon- in the initial (left-hand side) linear portion of the CO2 response curve and by biochemical processes at higher sive to CO2 concentration. Consequently, vertical CO2 concentrations. The CO2 compensation point is the variation in CO2 concentration within the canopy minimum CO2 concentration at which the leaf shows a net gain of carbon has relatively little effect on whole-ecosystem photosynthesis (Field 1991).

Although spatial variation in CO2 concentra- that reacts with CO2), and changes in stomatal tion does not explain much of the global variation opening have little influence on photosynthesis. in photosynthetic rate, the 35% increase in atmo-

At high internal CO2 concentrations, carboxyla- spheric CO2 concentration since the beginning of tion may be limited by: (1) insufficient light (or the industrial revolution has caused a general light-harvesting pigments) to provide energy, (2) increase in carbon gain by ecosystems (see Chap. 7; insufficient nitrogen invested in photosynthetic Canadell et al. 2007). In both growth-chamber enzymes to process the ATP, NADPH, and CO2 and field studies, a doubling of CO2 concentra- present in the chloroplast, or (3) insufficient tion increases photosynthetic rate by 30–50% phosphate or sugar phosphates to synthesize (Curtis and Wang 1998; Ainsworth and Long RuBP. 2005). This enhancement of photosynthesis by

Under a wide variety of circumstances, ter- elevated CO2 is most pronounced in C3 plants, restrial plants adjust the components of photo- especially woody species (Ainsworth and Long synthesis, so CO2 diffusion and biochemistry are 2005). Over time, most plants acclimate to ele- about equally limiting to photosynthesis vated CO2 by reducing photosynthetic capacity (Farquhar and Sharkey 1982), causing plants to and stomatal conductance, as expected from our respond to both CO2 availability and biochemi- hypothesis of co-limitation of photosynthesis by cal limitations (light, nitrogen, or phosphorus). biochemistry and diffusion. This down-regulation

Plants make this adjustment by altering stomatal of CO2 absorption in response to elevated CO2 conductance, which occurs within minutes, or by enables plants to sustain carbon uptake, while changing the concentrations of light-harvesting reducing transpiration rate and their water demand pigments or photosynthetic enzymes, which from soils. In this way, elevated CO2 often stimu- occurs over days to weeks. The general principle lates plant growth more strongly by reducing of co-limitation of photosynthesis by biochemis- moisture limitation than by its direct effects on try and diffusion provides the basis for under- photosynthesis. C4 plants are often just as sensitive standing most of the adjustments by individual to the indirect effects of CO2 as are C3 plants, leaves to minimize the environmental limitations so the long-term effects of elevated CO2 on the of photosynthesis. Stomatal conductance is competitive balance between C3 and C4 plants are regulated, so photosynthesis usually occurs near difficult to predict (Mooney et al. 1999). 140 5 Carbon Inputs to Ecosystems

1200 Light Limitation Clouds Above Physical environment determines light inputs Sunfleck canopy to ecosystems, and leaf area governs the distribution of light within the canopy. Leaves experience large fluctuations (10- to 1,000-fold) in incident light due to changes in sun angle, cloudi- 0 11:00 11:05 11:10 ness, and the location of sunflecks (patches of Below canopy direct sunlight that penetrate a plant canopy; Fig. 5.11). The vertical distribution of leaf area, ) however, is the major factor governing the light − 1 1200 s Clouds environment of individual leaves. Light distribu- − 2 tion within terrestrial canopies is approximated Dawn Dusk by an empirical relationship identical to that observed in aquatic ecosystems:

-kLz 0 Iz = Ie o (5.3)

Irradiance ( µ mol m 0 12:00 24:00 where I is irradiance at height z (m) beneath the canopy surface, Io is the irradiance at the top of Frontal system the canopy, k is the extinction coefficient per unit 1200 leaf area, and L is the leaf area index (LAI; the projected leaf area per unit of ground area) above the point of measurement. The actual distribution of light through the canopy is more complex and depends on the balance of direct and diffuse radiation. LAI is a key parameter governing ecosys- 0 June July tem processes because it determines both the area Time that is potentially available to absorb light and the degree to which light is attenuated through the Fig. 5.11 Hypothetical time course of photosynthetically canopy. LAI is equivalent to the total upper sur- active radiation above and below the canopy of a temperate forest over minutes, hours, and months. Over the face area of all leaves per area of ground (or the course of a few minutes, light at the top of the canopy var- projected leaf area in the case of cylindrical nee- ies with cloudiness. Beneath the canopy, light also varies dle-like leaves). due to the presence or absence of sunflecks of direct irra- LAI varies widely among ecosystems but typ- diance, which can last tenths of seconds to minutes. 2 −2 During a day, changes in solar angle and passing clouds ically has values of 1–8 m leaf m ground for cause large changes in light. Convective activity often ecosystems with a closed canopy. The extinction increases cloudiness in the afternoon. During the growing coefficient is a constant that describes the expo- season, seasonal changes in the solar angle and the passage nential decrease in irradiance through a canopy. of frontal systems are the major causes of variation in light. Some times of year have greater frequency of cloud- It is low for vertically inclined or small leaves iness than others due to changes in directions of the pre- (e.g., 0.3–0.5 for grasses), allowing substantial vailing winds and the passage of frontal systems penetration of direct radiation into the canopy, but high for near-horizontal leaves (0.7–0.8). Clumping of leaves around stems, as in conifers, most of the available light. Irradiance at the and variable leaf angles is associated with inter- ground surface of a forest, for example, is often mediate values for k. Equation (5.3) indicates that only 1–2% of that at the top of the canopy, simi- light is distributed unevenly in an ecosystem and lar to the light available at the bottom of aquatic that the leaves near the top of the canopy capture euphotic zones (Fig. 5.6). Terrestrial Photosynthesis 141

The shape of the light-response curve of Total ecosystem photosynthesis in terrestrial plants is identical

) Species C (sun) to that of aquatic algae (Fig. 5.5). Under light- C − 1 limiting conditions, photosynthesis increases lin- s − 2 early with increasing light availability (constant Species B quantum yield or light-use efficiency). As the B (intermediate) light-harvesting capacity of chlorophyll becomes A Species A (shade) light saturated, photosynthesis reaches its maximum rate (photosynthetic capacity). At extremely 0 high light, photosynthesis may decline due to photo-oxidation of pigments and enzymes, just as in phytoplankton (Fig. 5.5). Net photosynthesis ( µ mol m In response to fluctuations in light availabil- −2 −1 ity over minutes to hours (Fig. 5.11), plants alter Irradiance (µmol m s ) stomatal conductance to adjust CO supply to 2 Fig. 5.12 Light response curves of net photosynthesis in meet the needs of carbon-fixation reactions plants adapted (or acclimated) to low, intermediate, and (Pearcy 1990; Chazdon and Pearcy 1991). high light. Horizontal arrows show the range of irradiance Stomatal conductance increases in high light, over which net photosynthesis is positive and responds linearly to irradiance for each species and for the ecosys- when CO demand is high, and decreases in low 2 tem as a whole. Acclimation increases the range of light light, when photosynthetic demand for CO2 is availability over which net photosynthesis responds linearly low. These stomatal adjustments result in a rela- to light, i.e., has a constant light-use efficiency tively constant CO2 concentration inside the leaf, as expected from our hypothesis of co- limitation of photosynthesis by biochemistry Plants can also produce shade leaves as a and diffusion. It allows plants to conserve water result of adaptation, the genetic adjustment by a under low light and to maximize CO2 absorption population to maximize performance in a partic- at high light. ular environment. Species that are adapted to Over longer time scales (days to months), high light and intolerant of shade typically have a plants respond to variations in light availability higher photosynthetic capacity per unit mass or by producing leaves with different photosyn- area than do shade-tolerant species, even when thetic properties. This physiological adjustment growing in the shade (Walters and Reich 1999). by an organism in response to a change in some The main disadvantage of the high protein and environmental parameter is known as acclima- photosynthetic rate typical of shade-intolerant tion. Leaves at the top of the canopy (sun leaves) species is that they also have a higher respiration have more cell layers, are thicker, and therefore rate, due to their higher protein content. Species have a higher photosynthetic capacity per unit that are adapted to low light and are tolerant of leaf area than do shade leaves produced under shade have a low photosynthetic capacity, but can low light (Terashima and Hikosaka 1995; photosynthesize at lower light levels than shade- Walters and Reich 1999). The respiration rate of intolerant species. In other words, they have a a tissue depends on its protein content (see low light compensation point. At the light com- Chap. 6), so the low photosynthetic capacity pensation point, leaf respiration completely off- and protein content of shade leaves are associ- sets photosynthetic carbon gain, resulting in zero ated with a lower respiration rate per unit area net photosynthesis (Fig. 5.5). A mature shaded than in sun leaves. For this reason, shade leaves leaf typically does not import carbon from the maintain a positive carbon balance (photosyn- rest of the plant, so the leaf senesces and dies if it thesis minus respiration) under lower light lev- falls below the light compensation point for a els than do sun leaves (Fig. 5.12). long time. This puts an upper limit on the leaf 142 5 Carbon Inputs to Ecosystems

area that an ecosystem can support, regardless of ) − 1 1000 how favorable the climate and supply of soil s − 1 resources may be. On average, the leaf-level light compensation point of shade-tolerant species is 100 about half of that of shade-intolerant species (Craine and Reich 2005). Variations in leaf angle also influence the effi- 10 ciency with which a plant canopy uses light. At high light, plants produce leaves that are steeply Net photosynthesis (nmol g 1 angled, so they absorb less light (see Chap. 4). 7 21 63 This is advantageous because it reduces the prob- Leaf nitrogen (mg g−1) ability of overheating or photo-oxidation of photosynthetic pigments at the top of the canopy. At Fig. 5.13 Relationship between leaf-nitrogen concentra- the same time, it allows more light to penetrate to tion and maximum photosynthetic capacity (photosynthetic rate measured under favorable conditions) for plants lower leaves. Leaves at the bottom of the canopy, from Earth’s major biomes. Circles and the solid regres- on the other hand, are more horizontal in orienta- sion line are for 11 species from 6 biomes using a common tion to maximize light capture and are produced methodology. Crosses and the dashed regression line are in an arrangement that minimizes overlap with data from the literature. Redrawn from Reich et al. (1997) other leaves of the plant (Craine 2009). Do differences in light availability explain the photosynthesis within the canopy, but it is only a differences among ecosystems in carbon gain? In minor factor explaining regional variations in midsummer, when plants of most ecosystems are annual carbon inputs to ecosystems (Fig. 5.2). photosynthetically active, the daily input of visible light is nearly as great in the Arctic as in the tropics but is spread over more hours and is more Nitrogen Limitation diffuse at high latitudes (Billings and Mooney and Photosynthetic Capacity 1968). The greater daily carbon gain in the tropics than at high latitudes is therefore unlikely to Vascular plant species differ 10 to 50-fold in be a simple function of the light available to drive their photosynthetic capacity. Photosynthetic photosynthesis. Neither can variation in light capacity is the photosynthetic rate per unit leaf availability due to cloudiness explain differences mass measured under favorable conditions of among ecosystems in energy capture. The most light, moisture, and temperature. It is a measure productive ecosystems on Earth, the tropical and of the carbon-gaining potential per unit of bio- temperate rainforests, have a high frequency of mass invested in leaves. Photosynthetic capacity cloudiness, whereas arid grasslands and deserts, correlates strongly with leaf nitrogen concentra- which are less cloudy and receive nearly 10-fold tion (Fig. 5.13; Field and Mooney 1986; Reich more light annually, are less productive. Seasonal et al. 1997, 1999; Wright et al. 2004) because and interannual variations in irradiance can, how- photosynthetic enzymes account for a large pro- ever, contribute to temporal variation in carbon portion of the nitrogen in leaves (Fig. 5.2). Many gain by ecosystems. Aerosols emitted by volcanic ecological factors can lead to a high leaf-nitrogen eruptions and fires, for example, can reduce solar concentration and therefore a high photosynthetic irradiance and photosynthesis over large areas in capacity. Plants growing in high-nitrogen soils, particular years. Similarly, photosynthesis (GPP) for example, have higher tissue nitrogen concen- of the Amazon rainforest is greater in the dry sea- trations and photosynthetic rates than do the same son than under the cloudy conditions of the wet species growing on less fertile soils. This accli- season (Saleska et al. 2007). In summary, light mation of plants to a high nitrogen supply contrib- availability strongly influences daily and seasonal utes to the high photosynthetic rates in agricultural patterns of carbon input and the distribution of fields and other ecosystems with a rapid nitrogen Terrestrial Photosynthesis 143 )

) 15 − 1 − 1 1000 s − 1 bc 100

10 ce gl 1 0

te sc tr df

tu 1 5 Net photosynthesis (nmol g co td ) 100 mo dc − 1

Maximum stomatal conductance (mm s 0 010203040 10 Leaf nitrogen (mg g−1) Leaf nitrogen (mg g Fig. 5.14 Relationship between leaf-nitrogen concentration and maximum stomatal conductance of plants from 1

Earth’s major biomes. Each point and its standard error ) − 1

represent a different biome: bc, broad-leafed crops; ce, g 1000 2 cereal crops; co, evergreen conifer forest; dc, deciduous conifer forest; df, tropical dry forest; gl, grassland; mo, monsoonal forest; sc, sclerophyllous shrub; sd, dry savanna; sw, wet savanna; tc, tropical tree crop; td, temperate deciduous broadleaved forest; te, temperate evergreen 10 0 broadleaved forest; tr, tropical wet forest; tu, herbaceous tundra. Redrawn from Schulze et al. (1994) Specific leaf area (c m 10 110 100 Leaf lifespan (months) turnover. Many species differ in their leaf-nitrogen Fig. 5.15 The effect of leaf life span on photosynthetic concentration, even when growing in the same capacity, leaf-nitrogen concentration, and specific leaf soils. Species adapted to productive habitats usu- area. Symbols as in Fig. 5.13. Redrawn from Reich et al. ally produce leaves that are short-lived and have (1997) high tissue-nitrogen concentrations and high photosynthetic rates. Nitrogen-fixing plants also typically have high leaf-nitrogen concentrations and Conversely, species with a low photosynthetic correspondingly high photosynthetic rates. In capacity conserve water as a result of their lower summary, regardless of the cause of variation stomatal conductance. in leaf-nitrogen concentration, there is always a There appears to be an unavoidable tradeoff strong positive correlation between leaf-nitrogen between traits that maximize photosynthetic concentration and photosynthetic capacity rate and traits that maximize leaf longevity (Fig. 5.13; Reich et al. 1997; Wright et al. 2004). (Fig. 5.15; Reich et al. 1997, 1999; Wright et al. Plants with a high photosynthetic capacity have 2004). Many plant species that grow in low-nutrient a high stomatal conductance, in the absence of environments produce long-lived leaves because environmental stress (Fig. 5.14), as expected from nutrients are insufficient to support rapid leaf our hypothesis of co-limitation of photosynthesis turnover (Chapin 1980; Craine 2009). Shade- by biochemistry and diffusion. This enables plants tolerant species also produce longer-lived leaves with a high photosynthetic capacity to absorb than do shade-intolerant species (Reich et al.

CO2 rapidly, despite high rates of water loss. 1999; Wright et al. 2004). Long-lived leaves 144 5 Carbon Inputs to Ecosystems )

typically have a low leaf-nitrogen concentration − 1 2.0 and a low photosynthetic capacity; they must week

therefore photosynthesize for a relatively long − 1 1 time to break even in their lifetime carbon budget 1.5 (Gulmon and Mooney 1986; Reich et al. 1997). 3 2 To survive, long-lived leaves must have enough structural rigidity to withstand drought and winter 1.0 4 desiccation. These structural requirements cause 5 leaves to be dense, i.e., to have a small surface 6 0.5 area per unit of biomass, termed specific leaf area (SLA). Long-lived leaves must also be well 7

defended against herbivores and pathogens, if Maximum relative growth rate (g g 0 0 5101520 25 30 they are to persist. This requires substantial allo- 2 1 Maximum photosynthetic rate (µmol m− s− ) cation to lignin, tannins, and other non-nitrogenous compounds that deter herbivores, but also con- Fig. 5.16 Relationship between maximum photosynthetic rate and maximum relative growth rate for major tribute to tissue mass and a low SLA. plant growth forms: (1) agricultural crop species, (2) her- The broad relationship among species with baceous sun species, (3) grasses and sedges, (4) summer respect to photosynthetic rate and leaf life span is deciduous trees, (5) evergreen and deciduous dwarf shrubs, similar in all biomes; a twofold decrease in leaf (6) herbaceous shade species and bulbs; (7) evergreen conifers. Redrawn from Schulze and Chapin (1987) life span gives rise to about a fivefold increase in photosynthetic capacity (Reich et al. 1999;

Wright et al. 2004). ) − 1 1000 Plants in productive environments produce s − 1 short-lived leaves with a high tissue-nitrogen concentration and a high photosynthetic capacity; this 100 allows a large carbon return per unit of biomass invested in leaves, if enough light is available.

These leaves have a high SLA, which maximizes 10 the quantity of leaf area displayed and the light captured per unit of leaf mass. The resulting high Net photosynthesis (nmol g rates of carbon gain support a high maximum 101 70 490 relative growth rate in the absence of environmen- Specific leaf area (cm2 g−1) tal stress or competition from other plants (Fig. 5.16; Schulze and Chapin 1987). Many early Fig. 5.17 The relationship between specific leaf area (SLA) and photosynthetic capacity. The consistency of successional habitats, such as recently abandoned this relationship makes it possible to use SLA as an easily agricultural fields, canopy gaps, or post-fire sites, measured index of photosynthetic capacity. Symbols as in have enough light, water, and nutrients to support Fig. 5.13. Redrawn from Reich et al. (1997) high growth rates and are characterized by species with short-lived leaves, high tissue-nitrogen con- In summary, plants produce leaves with a concentration, high SLA, and high photosynthetic tinuum of photosynthetic characteristics, ranging rate (see Chap. 12). Even in late succession, envi- from short-lived, low-density leaves with a high ronments with high water and nutrient availability nitrogen concentration and high photosynthetic are characterized by canopy species with rela- rate to long-lived, dense leaves with a low nitro- tively high nitrogen concentration and photosyn- gen concentration and low photosynthetic rate. thetic rate. Plants in the canopy of these habitats These correlations among traits are so consistent can grow quickly to replace leaves removed by that SLA is often used in ecosystem comparisons herbivores or to fill canopy gaps produced by as an easily measured index of photosynthetic death of branches or individuals. capacity (Fig. 5.17). Terrestrial Photosynthesis 145

There is only modest variation in photosyn- loss by transpiration (see Chap. 4). The high-light thetic capacity per unit leaf area because conditions in which a plant would be expected to leaves with a high photosynthetic capacity per increase stomatal conductance to minimize CO2 unit leaf biomass also have a high SLA. limitations to photosynthesis are therefore often Photosynthetic capacity or assimilation rate per the same conditions in which the resulting transpi- unit leaf area (Aarea) is a measure of the capacity rational water loss is greatest and most detrimen- of leaves to capture a unit of incoming radiation. tal to the plant. This tradeoff between a response It is calculated by dividing photosynthetic (assim- that maximizes carbon gain (stomata open) and ilation) rate per unit leaf mass (Amass) by SLA. one that minimizes water loss (stomata closed) is typical of the physiological compromises Amass Aarea = (5.4) faced by plants whose physiology and growth SLA may be limited by more than one environmental --21 -- 11 2- 1 resource (Mooney 1972). When water supply is (gcm s )= (gg s ) / (cm g ) abundant, leaves typically open their stomata in response to high light, despite the associated There is relatively little variation in A among area high rate of water loss. As leaf water stress plants from different ecosystems (Lambers and develops, stomatal conductance declines to reduce Poorter 1992). In productive habitats, both mass- water loss (see Fig. 4.17). This decline in stomatal based photosynthesis and SLA are high conductance reduces photosynthetic rate and the (Fig. 5.15). In unproductive habitats, both of efficiency of using light to fix carbon (i.e., light- these parameters are low, resulting in modest use efficiency [LUE]) below levels found in variation in area-based photosynthetic rate unstressed plants. (Lambers and Poorter 1992). To the extent that Plant acclimation and adaptation to low water A varies among plants, it tends to be higher in area is qualitatively different than adaptation to low species with short-lived leaves (Reich et al. 1997). nutrients (Killingbeck and Whitford 1996; Mass-based photosynthetic capacity is a good mea- Cunningham et al. 1999; Wright et al. 2001; sure of the physiological potential for photosyn- Craine 2009). Plants in dry habitats typically thesis (the photosynthetic rate per unit of biomass have thicker leaves, similar leaf-nitrogen concen- invested in leaves). Area-based photosynthetic tration, and therefore more nitrogen per unit leaf capacity is a good measure of the efficiency of area than do plants in moist habitats. Dry-site these leaves at the ecosystem scale (photosynthetic plants also have a low stomatal conductance. This rate per unit of available light). Variation in soil combination of traits enables dry-site plants to resources has a much greater effect on the quantity maintain higher rates of photosynthesis at a given of leaf area produced than on the photosynthetic rate of water loss compared to plants in moist capacity per unit leaf area. sites (Cunningham et al. 1999; Wright et al. 2001). Dry-site leaves basically service more photosynthetic cells and photosynthetic capacity Water Limitation for a given stomatal conductance. Plants in dry areas minimize water stress by Water limitation reduces the capacity of indi- reducing leaf area (by shedding leaves or produc- vidual leaves to match CO2 supply with light ing fewer new leaves). Some drought-adapted availability. Water stress is often associated with plants produce leaves that minimize radiation high light because sunny conditions correlate absorption; their leaves reflect most incoming with low precipitation (low water supply) and radiation or are steeply inclined toward the sun with low humidity (high rate of water loss). High (see Chap. 4; Ehleringer and Mooney 1978). light also leads to an increase in leaf temperature High radiation absorption is a disadvantage in and water vapor concentration inside the leaf and dry environments because it increases leaf tem- therefore greater vapor pressure deficit and water perature, which increases respiratory carbon loss 146 5 Carbon Inputs to Ecosystems

Leaf

Mesophyll cells

Guard cell Stoma Epidermal cell

Boundary Cuticle layer

Bulk air CO2 H2O

Fig. 5.18 Cross section of a leaf, showing the diffusion pathways of CO2 and H2O into and out of the leaf, respectively. Length of the horizontal arrows outside the leaf is proportional to wind speeds in the boundary layer

(see Chap. 6) and transpirational water loss (see additional resistance to CO2 movement into the Chap. 4). Thus plants in dry environments have leaf, any change in stomatal conductance has a several mechanisms by which they reduce radia- proportionately greater effect on water loss than tion absorption to conserve water and carbon. on carbon gain. In addition, water diffuses more

The low leaf area, the reflective nature of leaves, rapidly than does CO2 because of its smaller and the steep angle of leaves are the main factors molecular mass and the steeper concentration accounting for the low absorption of radiation gradient that drives diffusion across the stomata. and low carbon inputs in dry environments. In For all these reasons, as stomata close, water loss other words, plants adjust to dry environments declines to a greater extent than does CO2 absorp- primarily by altering leaf area and radiation tion. The low stomatal conductance of plants in absorption rather than by reducing photosynthetic dry environments results in less photosynthesis capacity per unit leaf area. per unit of time but greater carbon gain per unit Water-use efficiency (WUE) of photosynthe- of water loss, i.e., greater WUE. Plants in dry sis is defined as the carbon gain per unit of water environments also enhance WUE by maintaining lost. WUE is quite sensitive to the size of sto- a somewhat higher photosynthetic capacity than matal openings because stomatal conductance would be expected for their stomatal conduc- has slightly different effects on the rates of CO2 tance, thereby drawing down the internal CO2 entry and water loss. Water leaving the leaf concentration and maximizing the diffusion gra- encounters two resistances to flow: the stomata dient for CO2 entering the leaf (Wright et al. and the boundary layer of still air on the leaf sur- 2001). Carbon isotope ratios in plants provide an face (Fig. 5.18). Resistance to CO2 diffusion from integrated index of WUE during plant growth the bulk air to the site of photosynthesis includes because the 13C concentration of newly fixed car- the same stomatal and boundary layer resistances bon increases under conditions of low internal plus an additional internal resistance associated CO2 concentration (Box 5.1; Ehleringer 1993). with diffusion of CO2 from the cell surface into C4 and CAM photosynthesis are additional adap- the chloroplast and any biochemical limitations tations that augment the WUE of plants, and ulti- associated with carboxylation. Because of this mately ecosystems. Terrestrial Photosynthesis 147

100 nitrogen and photosynthetic enzymes, which enable carboxylation to keep pace with the energy 80 supply from the light-harvesting reactions (Berry 60 and Björkman 1980). This explains why arctic Neuropogon 40 Atriplex and alpine plants typically have high leaf-nitrogen Tidestromia Ambrosia concentrations despite low soil-nitrogen availabil- 20 ity (Körner and Larcher 1988). Plants in cold Antarctic Desert Coastal dune 0 environments also have hairs and other morpho- −50 510152025303540 Photosynthesis (% of maximum) logical traits that raise leaf temperature above air Tissue temperature (C) temperature (Körner 1999). In hot environments Fig. 5.19 Temperature response of photosynthesis in with an adequate water supply, plants produce plants from contrasting temperature regimes. Species leaves with high photosynthetic rates. The associ- include antarctic lichen (Neuropogon acromelanus), a cool ated high transpiration rate cools the leaf, often coastal dune plant (Ambrosia chamissonis), an evergreen desert shrub (Atriplex hymenelytra), and a summer-active reducing leaf temperature below air temperature. desert perennial (Tidestromia oblongifolia). Redrawn from In hot, dry environments, plants close stomata Mooney (1986) to conserve water, and the cooling effect of transpiration is reduced. Plants in these environments often produce small leaves, which shed heat Temperature Effects effectively and maintain temperatures close to air temperature (see Chap. 4). In summary, despite Extreme temperatures limit carbon absorption. the sensitivity of photosynthesis to short-term Photosynthetic rate is typically highest near leaf variation in temperature, leaf properties minimize temperatures commonly experienced on sunny the differences in leaf temperature among eco- days (Fig. 5.19). Leaf temperature may differ sub- systems, and plants acclimate and adapt so there stantially from air temperature due to the cooling is no clear relationship between temperature and effects of transpiration, the effects of leaf surface average photosynthetic rate of leaves in the field, properties on energy absorption, and the influence when ecosystems are compared. of adjacent surfaces on the thermal and radiation environment of the leaf (see Chap. 4). At low temperatures, photosynthesis is limited directly by Pollutants temperature, as are all chemical reactions. At high temperatures, photosynthesis also declines, due to Pollutants reduce carbon gain, primarily by increased photorespiration and, under extreme reducing leaf area or photosynthetic capacity. conditions, enzyme inactivation and destruction of Many pollutants, such as SO2 and ozone, reduce photosynthetic pigments. Temperature extremes photosynthesis through their effects on growth often have a greater effect on photosynthesis than and the production of leaf area. Pollutants also does average temperature because of damage to directly reduce photosynthesis by entering the photosynthetic machinery (Berry and Björkman stomata and damaging the photosynthetic machin- 1980; Waring and Running 2007). ery, thereby reducing photosynthetic capacity Several factors minimize the sensitivity of (Winner et al. 1985). Plants then reduce stomatal photosynthesis to temperature. The enzymatically conductance to balance CO2 absorption with the controlled carbon-fixation reactions are typically reduced capacity for carbon fixation. This reduces more sensitive to low temperature than are the the entry of pollutants into the leaf, reducing the biophysically controlled light-harvesting reac- vulnerability of the leaf to further injury. Plants tions. Carbon-fixation reactions therefore tend to growing in low-fertility or dry conditions are limit photosynthesis at low temperature. Plants pre-adapted to pollutant stress because their low adapted to cold climates compensate for this by stomatal conductance minimizes the quantity of producing leaves with high concentrations of leaf pollutants entering leaves. Pollutants therefore 148 5 Carbon Inputs to Ecosystems affect these plants less than they affect rapidly leaves decreases exponentially through the canopy growing crops and other plants with high stomatal in parallel with the exponential decline in irradi- conductance. ance (Eq. (5.3); Hirose and Werger 1987). This is radically different from aquatic ecosystems, where turbulence causes regular mixing of the algal cells Terrestrial GPP in surface waters, and algae at all depths have a low photosynthetic capacity typical of shade GPP of terrestrial ecosystems integrates the plants. The matching of photosynthetic capacity to effects of environmental factors and leaf photo- light availability in terrestrial ecosystems is the synthetic properties through the canopy. GPP is response we expect from individual leaves within the sum of the net photosynthesis by all photosyn- the canopy because it maintains the co-limitation thetic tissue measured at the ecosystem scale. The of photosynthesis by diffusion and biochemical controls over GPP in terrestrial ecosystems are processes in each leaf. The matching of photo- more complex than in aquatic systems for at least synthetic capacity to light availability occurs three reasons: (1) Unlike aquatic systems, both the through the preferential transfer of nitrogen to quantity and photosynthetic properties of terres- leaves at the top of the canopy. At least three pro- trial photosynthetic tissues change from the top to cesses cause this to occur. (1) New leaves are the bottom of the canopy. (2) In addition to light produced primarily at the top of the canopy where and nutrients, which influence photosynthesis in light availability is highest, causing nitrogen to all ecosystems, terrestrial photosynthesis is sensi- be transported to the top of the canopy (Field tive to the availability of water and the delivery of 1983; Hirose and Werger 1987). (2) Leaves at the

CO2 to photosynthetic cells. (3) The structure of bottom of the canopy senesce when they become the plant canopy influences the delivery of light shaded below their light compensation point. and CO2 to, and the loss of water from, photosyn- Much of the nitrogen resorbed from these senesc- thetic cells. Despite these complexities, recent ing leaves (see Chap. 8) is transported to the top of technological developments allow measurement the canopy to support the production of young of fluxes of CO2 and other compounds at scales of leaves with high photosynthetic capacity. (3) Sun tens to thousands of square meters, making it pos- leaves at the top of the canopy develop more sible to measure whole-ecosystem carbon fluxes cell layers than shade leaves and therefore contain even in large-statured ecosystems like forests more nitrogen per unit leaf area. The accumulation (Baldocchi 2003). These measurements, when of nitrogen at the top of the canopy is most pro- combined with simulation modeling, permit esti- nounced in dense canopies, which develop under mation of GPP and other ecosystem carbon fluxes circumstances of high water and nitrogen avail- (see Box 7.2). In this chapter, we focus on ecologi- ability (Field 1991). In environments where leaf cal controls over GPP and consider its role in the area is limited by water, nitrogen, or time since ecosystem carbon balance in Chap. 7. disturbance, there is less advantage to concentrating nitrogen at the top of the canopy because light availability is high throughout the canopy. In these Canopy Processes sparse canopies, light availability, nitrogen concentrations, and photosynthetic rates show a more The vertical profile of leaf photosynthetic prop- uniform vertical distribution. erties in a canopy maximizes GPP in terrestrial Canopy-scale relationships between light and ecosystems. In contrast to pelagic ecosystems, nitrogen occur even in multi-species communi- leaves in terrestrial canopies remain fixed in the ties. In a single individual, there is an obvious same vertical location throughout their lives. Their selective advantage to optimizing nitrogen photosynthetic properties are therefore adapted distribution within the canopy because this pro- and acclimated to the environment where they are vides the greatest carbon return per unit of nitro- situated. In most closed-canopy ecosystems, for gen invested in leaves. We know less about the example, photosynthetic capacity of individual factors governing carbon gain in multi-species Terrestrial GPP 149

Jarvis 1991). For this reason, gas exchange in rough canopies is more tightly coupled to conditions in the free atmosphere than in smooth canopies.

t d Wind speed is important because it reduces the h e ig e L p thickness of the boundary layer of still air around s

d each leaf, producing steeper gradients in tempera- n Leaf area i

W ture and in concentrations of CO and water vapor

Tree height 2

e r

u from the leaf surface to the atmosphere. This

speeds the diffusion of CO into the leaf and the e 2 p

m loss of water from the leaf, enhancing both photo- e T synthesis and transpiration. A reduction in thick-

Fig. 5.20 Typical vertical gradients in leaf area, temper- ness of the leaf boundary layer also brings leaf ature, light, and wind speed in a forest. Temperature is temperature closer to air temperature. The net highest in the mid-canopy where most energy is absorbed. effect of wind on photosynthesis is generally pos- Based on Landsberg and Gower (1997) itive at moderate wind speeds and adequate moisture supply, enhancing photosynthesis at the top stands. In such stands, the individuals at the top of the canopy, where wind speed is highest. When of the canopy account for most of the photosyn- low soil moisture or a long pathway for water thesis and may be able to support greater root transport from the soil to the top of the canopy biomass to acquire more nitrogen, compared to reduces water supply to the uppermost leaves, as smaller subcanopy or understory individuals. in tall forests, the uppermost leaves reduce their This specialization and competition among indi- stomatal conductance, causing the zone of maxi- viduals probably contributes to the vertical scal- mum photosynthesis to shift farther down in the ing of nitrogen and photosynthesis observed in canopy. Although multiple environmental gradi- multi-species stands (Craine 2009). ents within the canopy have complex effects on Vertical gradients in other environmental photosynthesis, they probably enhance photosyn- variables often reinforce the maximization of thesis near the top of canopies in those ecosys- carbon gain near the top of the canopy. The tems with enough water and nutrients to develop canopy modifies not only light availability but also dense canopies. Variations in light and water other variables that influence photosynthetic rate, availability and leaf-nitrogen concentrations then including wind speed, temperature, relative humid- cause diurnal and seasonal shifts the height of ity, and CO2 concentration (Fig. 5.20). The most maximum photosynthesis within the canopy. important of these effects is the decrease in wind Canopy properties extend the range of light speed from the free atmosphere to the ground sur- availability over which the light-use efficiency face. The friction of air moving across Earth’s (LUE) of the canopy remains constant. The surface causes wind speed to decrease exponen- light-response curve of canopy photosynthesis, tially from the free atmosphere to the top of the measured in closed canopies (LAI > » 3), satu- canopy. In other words, Earth’s surface creates a rates at higher irradiance than does photosynthe- boundary layer similar to that which develops sis by a single leaf (Fig. 5.21) for several reasons around individual leaves (Fig. 5.18). Wind speed (Jarvis and Leverenz 1983). The more vertical continues to decrease from the top of the canopy angle of leaves in the upper canopy reduces the to the ground surface in ways that depend on can- probability of their becoming light saturated and opy structure. Smooth canopies, characteristic of increases light penetration into the canopy. The crops or grasslands, show a gradual decrease in clumped distribution of leaves in shoots, branches, wind speed from the top of the canopy to the and crowns also increases light penetration into ground surface, whereas rough canopies, charac- the canopy. Conifer canopies are particularly teristic of many forests, create more friction and effective in distributing light through the canopy turbulence that increases the vertical mixing of air due to the clumping of needles around stems. within the canopy (see Chap. 4; McNaughton and This could explain why conifer forests often 150 5 Carbon Inputs to Ecosystems

40 fact, at high light (and correspondingly high temperature and vapor pressure deficit), photosyn-

) thesis may decline in the upper canopy, causing

− 1 30

s shaded leaves to account for most of the total − 2 canopy photosynthesis under some circumstances y 20 p (Fig. 5.22; Law et al. 2002). o n a C In most ecosystems, including all forests that have been measured, GPP approaches a plateau 10 at high light, indicating a decline in LUE at high Leaf light (Fig. 5.23; Ruimy et al. 1995; Law et al. 2002; Turner et al. 2003b). This decline in LUE 0 Net photosynthesis (µ mol m at high light is most pronounced in low-resource environments with sparse canopies, where can-

−10 opy photosynthetic capacity is low, and all leaves 0 500 1000 1500 experience a similar light regime (Gower et al. Irradiance (µmol m−2 s−1) 1999; Baldocchi and Amthor 2001; Turner et al. 2003b). In other words, canopy photosynthetic Fig. 5.21 Light response curve of photosynthesis in a single leaf and a forest canopy. Canopies maintain a rela- response to light mirrors a photosynthetic tively constant LUE (linear response of photosynthesis to response that is similar to that of all individual light) over a broader range of light availability than do leaves. In dense canopies, more leaves are shaded individual leaves. Redrawn from Ruimy et al. (1995) and operate in the linear portion of the light- response curve, increasing LUE of the canopy as a whole (Fig. 5.23; Teskey et al. 1995; Turner 20 et al. 2003b).

) 15 − 1

s Leaf Area − 2

m l

2 ta o 10 T Variation in soil resource supply accounts for much of the spatial variation in leaf area and ves lea nlit Su GPP among ecosystem types. Analysis of satellite imagery shows that about 70% of the ice-free ter- GPP ( µ mol CO 5 restrial surface has relatively open canopies (LAI < 1; Fig. 5.24; Graetz 1991). GPP correlates Shaded leaves closely with leaf area below an LAI of about 4 0 0 500 1000 1500 2000 (Schulze et al. 1994), suggesting that leaf area is a critical determinant of GPP on most of Earth’s ter- Irradiance ( mol m−2 s−1) µ restrial surface, just as algal biomass or chloro- Fig. 5.22 Light response curve of GPP of a deciduous phyll is a key determinant of pelagic GPP (Fig. 5.1). forest, showing the contributions of shaded and sunlit GPP is less sensitive to LAI in dense canopies leaves, as calculated by the model CANVEG. Redrawn because the leaves in the middle and bottom of the from Law et al. (2002) canopy contribute relatively little to GPP over the course of a day or year. The availability of soil support a higher LAI than deciduous forests. The resources, especially water and nutrient supply, is light compensation point also decreases from the a critical determinant of LAI for two reasons: (1) top to the bottom of the canopy (Fig. 5.12), so Plants in high-resource environments produce a lower leaves maintain a positive carbon balance, large amount of leaf biomass, and (2) leaves pro- despite their relatively low light availability. In duced in these environments have a high SLA, i.e., Terrestrial GPP 151

20 30

) 25 − 1

d 15

− 2 20

10 15

GPP (g C m 5 5

0 0 2 4 6810 12 14 0 2 468 10 12 14

APAR (MJ m−2 d−1)

Fig. 5.23 Response of GPP to absorbed photosyntheti- of full sun, although there is considerable variability. The cally active radiation (APAR) in a Massachusetts deciduous grassland maintains a constant light-use efficiency over the forest (left) and a Kansas grassland (right). The forest main- entire range of naturally occurring irradiance. Redrawn tains a relatively constant light-use efficiency up to 30–50% from Turner et al. (2003b)

Open Closed (Fig. 5.17; Lambers and Poorter 1992; Reich et al. canopies (70%) canopies (30%) 1997; Wright et al. 2004). 50 Tropical Disturbances, herbivory, and pathogens forests (12%) reduce leaf area below levels that resources can 30 Boreal forest (7%) support. Soil resources and light extinction through Temperate the canopy determine the upper limit to the leaf forest (7%) area that an ecosystem can support. However, many 10 factors regularly reduce leaf area below this poten- Savanna tial LAI. Drought and freezing are climatic factors (18%) that cause plants to shed leaves. Other causes of leaf loss include physical disturbances (e.g., fire and 2 wind) and biotic agents (e.g., herbivores and patho- Grass/Shrublands gens). After major disturbances, the remaining (12%) plants may be too small, have too few meristems, or Crops lack the productive potential to quickly produce the Tundra (12%) Height of tallest stratum (m) (7%) leaf area that could potentially be supported by the Desert (27%) climate and soil resources of a site. For this reason, 0.25 LAI tends to increase with time after disturbance to an asymptote, then (at least in forests) often declines in late succession (see Chap. 12). 0.1 0.3 0.5 0.7 1.0 2.5 6.0 10.0 Human activities increasingly affect the leaf Leaf area index of tallest stratum area of ecosystems in ways that cannot be predicted [log scale] from climate. Overgrazing by cattle, sheep, and goats, for example, directly removes leaf area and Fig. 5.24 LAI and canopy height of the major biomes. causes shifts to vegetation types that are less pro- Typical values for that biome and the percentage of the terrestrial surface that it occupies are shown. The vertical ductive and have less leaf area than would other- line shows 100% canopy cover (LAI = 1). Redrawn from wise occur in that climate zone (Reynolds and Graetz (1991) Stafford Smith 2002). Acid rain and other pollutants can also cause leaf loss. Nitrogen deposition a large leaf area per unit of leaf biomass. As dis- can stimulate leaf production above levels that cussed earlier, a high SLA maximizes light capture would be predicted from climate and soil type, just and therefore carbon gain per unit of leaf biomass as nutrient and water additions to agricultural fields 152 5 Carbon Inputs to Ecosystems augment LAI and therefore GPP. Because of human to gain carbon (Xiao et al. 2010). In tropical activities, LAI cannot be estimated simply from ecosystems, however, where conditions are more correlations with climate. Fortunately, satellites continuously favorable for photosynthesis, leaves provide the opportunity to estimate LAI directly, maintain their photosynthetic machinery from the although the technology is still improving. Satellites time they are fully expanded until they are shed. tend to underestimate the LAI of dense canopies Models that simulate GPP often define the length because they cannot “see” all the leaves. LIDAR of the photosynthetic season in terms of thresholds (Light Detection and Ranging) uses reflection of of minimum temperature or moisture below which light pulses (lasers) to detect three-dimensional plants do not produce leaves or do not photosyn- canopy structure, much like radar, and shows prom- thesize (Running et al. 2004). ise in improving remote-sensing estimates of LAI. Environmental controls over GPP during Fortunately, most of the world’s canopies are rela- the growing season are similar to those described tively open, so their LAI can be estimated relatively for net photosynthesis of individual leaves. Soil accurately from satellites. Information about global resources (nutrients and moisture) influence GPP distribution of LAI is an important input to models primarily through their effects on photosynthetic that calculate regional patterns of carbon input to potential and leaf area rather than through varia- terrestrial ecosystems (Running et al. 2004). tions in the efficiency of converting light to carbohydrates (Turner et al. 2003b). Consequently, ecosystem differences in GPP depend more strongly Length of the Photosynthetic Season on differences in the quantity of light absorbed and length of photosynthetic season than on the efficiency The length of the photosynthetic season of converting light to carbohydrates (i.e., LUE). accounts for much of the ecosystem differences The seasonal changes in GPP depend on both in GPP. Most ecosystems experience times that the seasonal patterns of leaf area development and are too cold or too dry for significant photosynthe- loss and the photosynthetic response of individual sis to occur. During winter in cold climates and leaves to variations in light and temperature, which times with negligible soil water in dry climates, influence LUE. These environmental factors have plants either die (annuals), lose their leaves (decid- a particularly strong effect on leaves at the top of uous plants), or become physiologically dormant the canopy, which account for most GPP. The thin- (some evergreen plants). During these times, there ner boundary layer and greater distance for water is negligible carbon absorption by the ecosystem, transport from roots, for example, makes the upper- regardless of light availability and CO2 concentra- most leaves particularly sensitive to variation in tion. In a sense, the non-photosynthetic season is temperature, soil moisture, and relative humidity. simply a case of extreme environmental stress. At LUE varies diurnally, being lowest at times of high latitudes and altitudes and in dry ecosystems, high light. Seasonal patterns of LUE are more this is probably the major constraint on carbon complex because they depend not only on light inputs to ecosystems (Fig. 5.2; see Chap. 6; Körner availability but also on seasonal variations in leaf 1999). For annuals and deciduous plants, the lack area, canopy nitrogen, and various environmental of leaf area is sufficient to explain the absence of stresses such as drought and freezing. LUE is photosynthetic carbon gain in the nongrowing highest in high-resource ecosystems such as crops season. Lack of water or extremely low tempera- with a high LAI and photosynthetic capacity. LUE tures can, however, prevent even evergreen plants is lowest in low-resource ecosystems such as the from gaining carbon. Some evergreen species par- boreal forest and arid grasslands (Turner et al. tially disassemble their photosynthetic machinery 2003b). LUE also declines with increasing tem- during the nongrowing season. These plants perature (reflecting increases in photorespiration; require some time after the return of favorable Lafont et al. 2002; Turner et al. 2003b) and is environmental conditions to reassemble their pho- strongly reduced at extremely low temperatures tosynthetic machinery (Bergh and Linder 1999), (Teskey et al. 1995). The detailed patterns and so not all early-season irradiance is used efficiently causes of temporal and spatial patterns of LUE Terrestrial GPP 153 and GPP are active research areas that promise to lenge, however, is to estimate the fraction of provide important advances in understanding and absorbed radiation that has been absorbed by leaves predicting patterns of carbon inputs to ecosystems rather than by soil or other non-photosynthetic (Running et al. 2004; Luyssaert et al. 2007; surfaces. Vegetation has a different spectrum of Waring and Running 2007). absorbed and reflected radiation than does the atmosphere, water, clouds, or bare soil. This occurs because chlorophyll and associated light-harvesting Satellite-Based Estimates of GPP pigments or accessory pigments, which are con- centrated at the canopy surface, absorb visible Satellite-based estimates of absorbed radiation light (VIS) efficiently. The optical properties that and LUE allow daily mapping of GPP at global result from the cellular structure of leaves, how- scales. An important conclusion of leaf- and can- ever, make them highly reflective in the near infra- opy-level studies of photosynthesis is that many red (NIR) range. Ecologists have used these unique factors cause convergence of ecosystems toward a properties of vegetation to generate an index of relatively similar efficiency of converting absorbed vegetation “greenness”: the normalized differ- light energy into carbohydrates. (1) All C3 plants ence vegetation index (NDVI). have a similar quantum yield (LUE) at low to mod- (NIR- VIS) erate irradiance. (2) Penetration of light and verti- NDVI = (5.5) cal variations in photosynthetic properties through (NIR+ VIS) a canopy extend the range of irradiance over which NDVI is approximately equal to the fraction of LUE remains relatively constant. (3) LUE of a incoming photosynthetically active radiation given ecosystem varies primarily in response to (PAR) that is absorbed by vegetation (FPAR): light intensity and short-term environmental stresses FPAR»» NDVI APAR / PAR (5.6) that reduce stomatal conductance. Over the long term, however, plants respond to environmental where APAR is the absorbed photosynthetically stresses by reducing leaf area and the concentra- active radiation (Running et al. 2004). FPAR can tions of photosynthetic pigments and enzymes so also be measured directly in ecosystems, know- photosynthetic capacity matches stomatal conduc- ing the irradiance at the top (Io) and bottom (Iz) of tance. In other words, plants in low-resource envi- the canopy or the relationship between Io and leaf ronments reduce the amount of light absorbed more area index (LAI, L): strongly than they reduce the efficiency with which FPAR=- 1 (II / ) (5.7) absorbed light is converted to carbohydrates. zo −kLz Modeling studies and field measurements suggest where Iz = I o e , and k is the extinction coeffi- that ecosystems differ much more strongly in leaf cient (5.3). Sites with a high rate of carbon gain area and photosynthetic capacity than in LUE generally have a high NDVI because of their high (Field 1991; Turner et al. 2003b). chlorophyll content (low reflectance of VIS) and If LUE is indeed similar and shows predictable high leaf area (high reflectance of NIR). Species patterns among ecosystems, GPP can be estimated differences in leaf structure also influence infra- from satellite measurements of light absorption red reflectance (and therefore NDVI). Conifer for- by ecosystems, and correcting this for known ests, for example, generally have a lower NDVI causes of variation in LUE. Leaves at the top of than deciduous forests despite their greater leaf the canopy have a disproportionately large effect area. Consequently, NDVI must be used cau- on the light that is both absorbed and reflected by tiously when comparing ecosystems dominated the ecosystem. Satellites can measure the incom- by structurally different types of plants (Verbyla ing and reflected radiation. This similarity in bias 1995). The maximum NDVI measured by satel- between the vertical distribution of absorbed and lites is very similar to that measured on the ground reflected radiation makes satellites an ideal tool (Fig. 5.25). If LUE is known, GPP can be calcu- for estimating canopy photosynthesis. The chal- lated from irradiance (PAR) and FPAR or NDVI: GPP=´ LUE FPAR ´»´ PAR LUE NDVI ´ PAR (5.8) 154 5 Carbon Inputs to Ecosystems

MODIS (Moderate Resolution Imaging have estimated LUE for different biomes, under Spectroradiometer) sensors carried aboard satel- varying conditions of vapor pressure deficit and lites directly measure reflectance from space, temperature (Running et al. 2000; White et al. allowing calculation of NDVI. Ecosystem models 2000). Using these modeled LUE values (g carbon MJ−1) and observed climate, NDVI and PAR (MJ m−2), daily GPP (g carbon m−2) can now be 1.0 calculated globally at a 1-km scale (Running et al. 0.8 2004). These calculations are based on daily observations of weather, weekly estimates of 0.6 NDVI, and annual estimates of biome distribu- 0.4 tions. The methodology for estimating global patterns of GPP is continually being tested and Measured FPAR 0.2 improved. Currently, differences in the scale at which weather observations are made account for 0 0 0.2 0.4 0.6 0.8 1.0 much of the discrepancy between GPP estimates FPAR from NDVI from satellites and those measured at specific field sites. Other sources of variation include the con- Fig. 5.25 Relationship between FPAR (the fraction of trols over GPP that were described in the previous photosynthetically active radiation absorbed by vegeta- section (Turner et al. 2005; Heinsch et al. 2006). tion) estimated from satellite measurements of NDVI (X-axis) and FPAR measured in the field (Y-axis). Data In the conterminous U.S. summer, GPP is highest were collected from a wide range of ecosystems, includ- in fertile moist ecosystems like croplands and ing temperate and tropical grasslands and temperate and deciduous forests and lowest in dry ecosystems boreal conifer forests. Satellites provide an approximate like grasslands and forests (Fig. 5.26). Evergreen measure of the photosynthetically active radiation absorbed by vegetation and therefore the carbon inputs to forests have modest mid-summer GPP but ecosystems. Redrawn from Los et al. (2000) continue photosynthesizing during the winter.

15 All vegetation Evergreen forests (EF) Deciduous forest (DF) 12 Mixed forests (MF) Shrublands (Sh) ) Savannas (Sa) − 1

d Grasslands (Gr) − 2 9 Croplands (Cr)

6 GPP (g C m

0 Jan Mar Jun Sep Dec Time

Fig. 5.26 Predicted seasonal pattern of GPP in differ- (a network of ecosystem flux studies) GPP measure- ent biomes of the U.S. averaged from 2001 to 2006, ments and MODIS satellite imagery. Redrawn from based on a regression model that uses AmeriFlux Xiao et al. (2010) Review Questions 155

These seasonal and ecosystem differences in GPP are determined primarily by leaf area and sec- are the major factors explaining ecosystem differ- ondarily by environmental stresses that reduce ences in NPP (see Chap. 6) and carbon accumula- the efficiency with which these leaves convert tion (see Chap. 7). light to chemical energy.

Review Questions Summary

1. How do light, CO2, and nitrogen interact to Most carbon enters terrestrial ecosystems through influence the biochemistry of photosynthesis photosynthesis mediated by primary producers in C3 plants? What biochemical adjustments (plants on land and phytoplankton in aquatic eco- occur when each of these resources declines in systems). The light-harvesting reactions of pho- availability? tosynthesis transform light energy into chemical 2. Describe the environmental controls over pho- energy, which is used by the carbon-fixation reac- tosynthesis in pelagic ecosystems in terms of tions to convert CO2 to sugars. The enzymes that the photosynthetic response of individual cells carry out these reactions account for about half of (e.g., light response curve) and ecosystem- the nitrogen in photosynthetic cells. scale photosynthesis (GPP). In pelagic ecosystems, phytoplankton are rela- 3. How does each major environmental variable tively well mixed throughout the euphotic zone (CO2, light, nitrogen, water, temperature, pol- and have photosynthetic properties similar to lutants) affect photosynthetic rate in terrestrial shade plants. Aquatic GPP depends on the quan- plants in the short term? How do plants adjust tity of phytoplankton and the vertical profile of to changes in each factor over the long term? light and other physical factors. Nutrient avail- 4. How does the response of photosynthesis to ability, as affected by stratification and vertical one environmental variable (e.g., water or mixing, strongly influences phytoplankton abun- nitrogen) affect the response to other environ- dance and therefore GPP. mental variables (e.g., light, CO2, or pollut- Plants on land adjust the components of pho- ants)? Considering these interactions among tosynthesis so physical and biochemical pro- environmental variables, how might anthropo- cesses co-limit carbon fixation. At low light, for genic increases in nitrogen inputs affect the example, plants reduce the quantity of photo- response of Earth’s ecosystems to rising atmo- synthetic machinery per unit leaf area by pro- spheric CO2? ducing thinner leaves. As atmospheric CO2 5. How do environmental stresses affect light- concentration increases, plants reduce stomatal use efficiency in the short term? How does conductance. The major environmental factors vegetation adjust to maximize LUE in stress- that explain differences among ecosystems in ful environments over the long term? carbon gain are the length of time during which 6. What factors are most important in explaining conditions are suitable for photosynthesis and differences among ecosystems in GPP? Over the soil resources (water and nutrients) available what timescale does each of these factors have to support the production and maintenance of its greatest impact on GPP? Explain your leaf area. Environmental stresses, such as inad- answers. equate water supply, extreme temperatures, and 7. What factors most strongly affect leaf area pollutants, reduce the efficiency with which and photosynthetic capacity of vegetation? plants use light to gain carbon. Plants also 8. How do the factors regulating photosynthesis respond to these stresses by reducing leaf area in a forest canopy differ from those in indi- and nitrogen content so as to maintain a rela- vidual leaves? How do availability of soil tively constant efficiency in the use of light to resources (water and nutrients) and the struc- fix carbon. Consequently, ecosystem differences ture of the canopy influence the importance of in photosynthesis at the ecosystem scale (GPP) these canopy effects? 156 5 Carbon Inputs to Ecosystems

Schulze, E.-D., F. M. Kelliher, C. Körner, J. Lloyd, and Additional Reading R. Leuning. 1994. Relationship among maximum stomatal conductance, ecosystem surface conduc- Craine, J.M. 2009. Resource Strategies of Wild Plants. tance, carbon assimilation rate, and plant nitrogen Princeton University Press, Princeton. nutrition: A global ecology scaling exercise. Annual Lambers, H., F.S. Chapin, III, and T.L. Pons. 2008. Plant Review of Ecology and Systematics 25:629–660. Physiological Ecology. 2nd edition. Springer, New York. Turner, D.P., S. Urbanski, D. Bremer, S.C. Wofsy, Law, B.E., E. Falge, L. Gu, D.D. Baldocchi, P. Bakwin T. Meyers et al. 2003. A cross-biome comparison of et al. 2002. Environmental controls over carbon diox- daily light use efficiency for gross primary production. ide and water vapor exchange of terrestrial vegetation. Global Change Biology 9:383–395. Agricultural and Forest Meteorology 113:97–120. Waring, R.H. and S.W. Running. 2007. Forest Ecosystems: Running, S.W., R.R. Nemani, F.A. Heinsch, M. Zhao, Analysis at Multiple Scales. 3rd edition. Academic M. Reeves, and H. Hashimoto. 2004. A continuous Press, San Diego. satellite-derived measure of global terrestrial primary Wright, I.J., P.B. Reich, M. Westoby, D.D. Ackerly, production. BioScience 54:547–560. Z. Barusch et al. 2004. The world-wide leaf economics spectrum. Nature 428:821–827. Sage, R.F. 2004. The evolution of C4 photosynthesis. New Phytologist 161:341–370.