Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Structural properties related to mesophyll conductance and underlying variation in leaf mass area of… Milla-Moreno, Estefania 2015

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2015_may_millamoreno_estefania.pdf [ 3.37MB ]
Metadata
JSON: 24-1.0167666.json
JSON-LD: 24-1.0167666-ld.json
RDF/XML (Pretty): 24-1.0167666-rdf.xml
RDF/JSON: 24-1.0167666-rdf.json
Turtle: 24-1.0167666-turtle.txt
N-Triples: 24-1.0167666-rdf-ntriples.txt
Original Record: 24-1.0167666-source.json
Full Text
24-1.0167666-fulltext.txt
Citation
24-1.0167666.ris

Full Text

STRUCTURAL PROPERTIES RELATED TO MESOPHYLL CONDUCTANCE AND UNDERLYING VARIATION IN LEAF MASS AREA OF BALSAM POPLAR (Populus balsamifera L.) by Estefania Milla-Moreno B. Sc. in Forestry, University of Chile, 2010   A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE  in The Faculty of Graduate and Postdoctoral Studies  (Forestry) THE UNIVERSITY OF BRITISH COLUMBIA  (Vancouver)  February 2015 © Estefania Milla-Moreno, 2015  ii Abstract In leaves, the ease with which CO2 may diffuse from substomatal cavities to the sites of carboxylation (mesophyll conductance, gm) is inversely proportional to the pathway length in the gas phase and the structural resistances encountered in the liquid phase. Increased length of any pathway component should decrease gm, whereas increased area for diffusion should increase gm. There is evidence that within native balsam poplar (Populus balsamifera L.) populations, gm may increase with latitude of origin, as do leaf mass (LMA) and nitrogen per unit area. To investigate the internal structural characteristics that might limit maximum gm in leaves, a balsam poplar family (K4×C) known to have high variation in LMA was chosen. Several whole tissue properties (LMA, carbon-nitrogen ratio, chlorophyll concentration index), microanatomy traits (leaf thickness, intercellular air space, cell wall surface area, chloroplast counts), and ultrastructural attributes (cell wall thickness, thylakoid grana thickness, and cell wall area adjacent to chloroplasts) were measured. There were significant genotypic differences in chlorophyll concentration index, numbers of chloroplasts per mesophyll cell, leaf thickness (tleaf), and thicknesses and cross-sectional areas of the palisade and spongy mesophyll layers. There were also differences in the fraction of intercellular air space (fias), as well as total and exposed cell wall surface areas of the mesophyll and its component tissues. Although genotypic differences in LMA were not significant, LMA varied as a function of tleaf (r=0.515, p<0.05) and fias (r=-0.510, p<0.05), which together explained considerable variation in this trait. The single best correlate of LMA was the cell wall area of the palisade (r=0.813, p<0.001), which accounted for well over half of the total cell wall area of the mesophyll. The areas of  iii mesophyll cell walls and chloroplasts exposed to intercellular air space, both of which should contribute to gm, also increased with LMA. Along the same lines, but not correlated with LMA, there was a decrease in nitrogen density per unit exposed mesophyll surface area associated with thicker leaves (r=-0.481, p<0.05).       iv Preface Dr. Robert Guy and Dr. Raju Soolanayakanahally conceived the study. Dr. Robert Guy advised on data interpretation and helped edit the thesis. Dr. Lacey Samuels and Dr. Christopher Chanway advised on data interpretation. PhD candidate Mina Momayyezi shared her preliminary data on mesophyll conductance for the family used in my thesis. Dr. Anthony Kozak advised on statistics. Dr. Athena McKown provided advice on microscopy related questions, draft editing and also helped harvest the plant material, together with Jose Arias, Linda Quamme and Catherine Denny. I designed the experiment, took care of the plant material, harvested the plant material, standardized the microscopy and protocols, analyzed the data and drafted the thesis.     v Table of contents Abstract ........................................................................................................................................ ii Preface ......................................................................................................................................... iv Table of contents....................................................................................................................... v List of tables ............................................................................................................................ viii List of figures ............................................................................................................................. ix List of abbreviations ............................................................................................................... xi Acknowledgements ............................................................................................................... xiv Dedication ................................................................................................................................ xvi CHAPTER 1. Introduction ....................................................................................................... 1 1.1. AgCanBaP germplasm ............................................................................................................... 3 1.2. Linking anatomy with physiology ........................................................................................ 4 1.3. Leaf mass per unit area (LMA) ............................................................................................... 8 1.4. Nitrogen content (N) ................................................................................................................. 9 1.5. Connections between traits ................................................................................................. 10 1.6. Leaf anatomy in poplar .......................................................................................................... 11 1.7. Research hypothesis and objectives ................................................................................. 12 CHAPTER 2.  Material and methods ................................................................................ 14 2.1. Plant material ........................................................................................................................... 14 2.2. Developmental stage of the plants .................................................................................... 16 2.3. Leaf mass per unit area, carbon and nitrogen content (LMA, C%, N%) ............... 17 2.4. Chlorophyll concentration index (CCI) ............................................................................ 17  vi 2.5. Stomatal density (stoD) and length of the stomatal aperture (stoLA) .................. 18 2.6. Microscopy ................................................................................................................................. 18 2.6.1. Microanatomy by light microscopy ......................................................................................... 19 2.6.2. Ultrastructure by transmission electron microscopy (TEM)........................................ 21 2.7. Statistical analyses .................................................................................................................. 22 CHAPTER 3. Results .............................................................................................................. 24 3.1. Developmental stage of the plants .................................................................................... 24 3.1.1. Leaf plastochron index (LPI) ...................................................................................................... 24 3.2. Whole tissue properties ........................................................................................................ 25 3.2.1. Leaf mass area (LMA) .................................................................................................................... 25 3.2.2. Carbon-Nitrogen ratio (C:N) and Nitrogen density (Na) ................................................. 25 3.2.3. Chlorophyll concentration index (CCI) .................................................................................. 27 3.3. Microanatomy ........................................................................................................................... 28 3.3.1. Leaf thickness (tleaf) and its components ............................................................................... 29 3.3.2. Cross-sectional areas (Ap, Aupp, Alow, Aab and Aad) ................................................................ 29 3.3.3 Fraction of mesophyll occupied by intercellular air space (fias) ................................... 29 3.3.4. Surface area of mesophyll cells (Pmes/S) and mesophyll surface area exposed to intercellular air space per unit leaf area (Smes/S)........................................................................... 33 3.3.5. Chloroplasts in mesophyll tissues (Chlo_P, Chlo_US) ....................................................... 37 3.3.6 Nitrogen density per unit mesophyll wall area Na/Smes/S ............................................... 37 3.4. Ultrastructure ........................................................................................................................... 39 3.4.1. Cell wall thickness (tcw) ................................................................................................................ 40 3.4.2. Thylakoid grana thickness (tgra) ................................................................................................ 40 3.4.3. Chloroplast surface area exposed to intercellular air space (Sc/S) ............................ 41 3.5. Stomata ....................................................................................................................................... 42  vii 3.6. Trait correlations .................................................................................................................... 43 CHAPTER 4. Discussion ........................................................................................................ 48 4.1. Study limitations ...................................................................................................................... 48 4.2. Whole tissue properties ........................................................................................................ 50 4.3. Microanatomy ........................................................................................................................... 51 4.4. Ultrastructure ........................................................................................................................... 53 4.5. Stomata ....................................................................................................................................... 55 4.6. Structural basis of variation in LMA ................................................................................. 56 4.7. Correlation with gm ................................................................................................................. 56 4.8. Conclusions ................................................................................................................................ 57 Bibliography ............................................................................................................................ 59 Appendix: Pearson’s product-moment correlation coefficients (r) of gm on measured traits ...................................................................................................................... 65          viii List of tables Table 1. Geographic coordinates of the parental material ................................................ 14 Table 2. Analysis of variance of the leaf plastochron index (n=2 to 5) ............................ 24 Table 3. Analysis of variance of the chlorophyll concentration (n=2 to 5) ...................... 27 Table 4. Analysis of variance for leaf, palisade and spongy mesophyll thickness (n=2) .......................................................................................................................... 30 Table 5. Analysis of variance of cross-sectional cell area within mesophyll tissues, per mm of leaf surface (n=2) .................................................................................... 31 Table 6. Analysis of variance of cross-sectional area of adaxial and abaxial epidermal layers, per mm of leaf surface (n=2). ....................................................... 32 Table 7. Analysis of variance of fias values (n=2) ............................................................. 33 Table 8. Analysis of variance of Pmes/S_T and Smes/S_T values (n=2) .............................. 34 Table 9. Analysis of variance of Pmes/S_P, Pmes/S_US and Pmes/S _LS values (n=2) ....... 35 Table 10. Analysis of variance of Smes/S_P, Smes/S_US and Smes/S_LS values (n=2) ....... 36 Table 11. Analysis of variance for the count of chloroplasts, visible in cross-section, in palisade and upper spongy mesophyll (n=2) ........................................................ 37 Table 12. Analysis of variance for the stomatal count per mm2 (n=2-5) .......................... 43 Table 13. Pearson’s product-moment correlation coefficients (r). ................................... 44  ix List of figures Figure 1. The native range of Populus balsamifera, indicating locations sampled to construct the AgCanBaP collection ............................................................................ 4 Figure 2. Light micrograph of a Populus balsamifera leaf cross-section showing the diffusion of CO2 into leaves ........................................................................................ 5 Figure 3. Diffusion pathways into chloroplasts. a) Transmission electron micrograph of palisade cells demonstrating parallel diffusion pathways within cells determined by the exposure of chloroplasts to intercellular air space. b) Transmission electron micrograph of the chloroplast circled in green in the first panel ............................................................................................................................ 6 Figure 4. Development of the plants at the greenhouse over the first 8 weeks of growth (a to h, respectively). .................................................................................... 15 Figure 5. Leaf mass per unit area, by clone ...................................................................... 25 Figure 6. Carbon-Nitrogen ratio, by clone ........................................................................ 26 Figure 7. Nitrogen content per unit area, by clone ........................................................... 26 Figure 8. Light micrographs of K4C9 and K4C13 indicating thickness measurements within the leaf and morphological differences between the two clones. .................. 28 Figure 9. Nitrogen content per unit Smes/S_T, by clone .................................................... 38 Figure 10. Transmission electron micrograph of K4C1 with all the specifications. ........ 39  x Figure 11.  Cell wall thickness averages, by clone ........................................................... 40 Figure 12.  Thylakoid grana thickness averages, by clone ............................................... 41 Figure 13.  Chloroplast surface area exposed to intercellular air space per unit of leaf area, by clone ............................................................................................................ 41 Figure 14. Stomatal impressions for genotypes K4C14, K4C18 and K4C20 .................. 42 Figure 15. Genotypic variation in LMA as a function of both tleaf and fias ....................... 46    xi List of abbreviations Abbreviation Definition Unit Aad adaxial cross-sectional cell area (per mm width of leaf) mm2 mm-1 Aab abaxial cross-sectional cell area mm2 mm-1 AgCanBaP Agriculture Canada Balsam Poplar Collection Not appropriate Alow lower spongy mesophyll cross-sectional area, excluding intercellular air space mm2 mm-1 Ap palisade cross-sectional area, excluding intercellular air space mm2 mm-1 As spongy mesophyll cross-sectional area, excluding intercellular air space mm2  Aupp upper spongy mesophyll cross-sectional area, excluding intercellular air space mm2 mm-1 Ci,w CO2 concentration from the sub-stomatal space to the outer surface of cell walls mol CO2 m-2 s-1 Ci,w-Cc CO2 drawdown determined by gliq  CCI chlorophyll content index dimensionless C:N carbon-to-nitrogen ratio mg mg-1 Chlo_P count of visible chloroplasts per  palisade cells #/cell Chlo_US count of visible chloroplasts per  upper spongy mesophyll cells #/cell ELEV elevation  m fias fraction of mesophyll occupied by intercellular airspace dimensionless gliq liquid-phase diffusion conductance mol CO2 m-2 s-1 gm internal (mesophyll) conductance mol CO2 m-2 s-1 gs stomatal conductance mol H2O m-2 s-1 LAT latitude °N Lc' total length of the chloroplasts exposed to the intercellular air space mm  xii Abbreviation Definition Unit LMA leaf mass per unit area mg mm-2 Lmes total length of mesophyll cells facing the intercellular airspace mm Lmes’ total length of chloroplasts facing the intercellular airspace mm LONG longitude  °W Lp leaf perimeter cm LPI leaf plastochron index dimensionless Na nitrogen per unit leaf area mg mm2 Na/Smes/S nitrogen per unit mesophyll area mg mm2 Pmes/S surface area of mesophyll cells, per unit area of leaf surface, calculated from total perimeter of cell walls m2 m-2 Pmes/S_P surface area of mesophyll cells calculated from total perimeter of cell walls in palisade tissue m2 m-2 Pmes/S_US surface area of mesophyll cells calculated from total perimeter of cell walls in upper spongy mesophyll  m2 m-2 Pmes/S_LS surface area of mesophyll cells calculated from total perimeter of cell walls in lower spongy mesophyll m2 m-2 Pmes/S_T total surface area of mesophyll cells calculated from total perimeter of cell walls m2 m-2 PPFP Photosynthetic photon flux density μmol m-2 s-1 Sc/S  chloroplast surface area exposed to intercellular airspace per unit of leaf area m2 m-2 Smes/S  surface area of mesophyll cells exposed to airspace per unit of leaf area m2 m-2 Smes/S_P surface area of mesophyll cells exposed to airspace in palisade tissue m2 m-2 Smes/S_US surface area of mesophyll cells exposed to airspace in upper spongy mesophyll m2 m-2 Smes/S_LS surface area of mesophyll cells exposed to airspace in lower spongy mesophyll m2 m-2  xiii Abbreviation Definition Unit Smes/S_T total surface area of mesophyll cells exposed to airspace, per unit of leaf area m2 m-2 stoD stomatal density per mm2 stoLA length of the stomatal aperture μm tcw cell wall thickness nm tgra thylakoid grana thickness nm tleaf leaf thickness mm tmes mesophyll thickness between the two epidermal layers mm tpal palisade thickness mm tspo spongy mesophyll thickness mm W width of the section measured mm ∑Ss  cross-sectional area of the mesophyll cells mm2 γ curvature correction factor dimensionless     xiv Acknowledgements I want to give a heartfelt thanks to my supervisor Dr. Robert Guy for his confidence, which gave me independence to develop this research. At the same time, he provided me with important advice every time I needed it. He helped to solve any doubt or difficulty this research aroused. He has striven to build a friendly lab environment, which has made of this endeavor a much more pleasant experience.  I offer my gratitude as well to my supervisory committee, Dr. Lacey Samuels and Dr. Christopher Chanway, for their insights and important feedback during my research.  I owe particular thanks to Dr. Raju Soolanayakanahally, who was responsible for the framework of this work and for providing me the connection to my funding support from Agriculture Canada. He has been continuously supportive from the very beginning until the last stage. I also thank Natalie Ryan and Mina Momayyezi, whose work helped me to contextualize the frame of this research. I would like to acknowledge CONICYT-Becas Chile and Agriculture and Agri-food Canada -Science and Technology Branch, whose funding opportunities made this research possible. I extend my deep gratitude to Dan Naidu, Gayle Kosh and Cindy Prescott, UBC Forestry Faculty and Staff, for their continuing support in academic and non-academic matters.  Special thanks to my lab group, where I found good friends. Thanks to Shofi, Lee, Mina, Athena, Linda, Natalie, Richard, Catherine, Daniela and Limin whose help and advises are embedded in this work. I also want to recognize and thank the statistical support of Dr. Anthony Kozak. Thanks to the BioImaging staff: Derrick, Brad, Garnet and Kevin,  xv who trained me in all the microscopy techniques and solved all type of questions that the imaging procedure aroused.  Foremost, I deeply thank to my mom, my husband Jose and my daughters, for their continuous love, support and encouragement during my research.     xvi Dedication To my treasured husband and daughters         1 CHAPTER 1. Introduction “Leaves bring the plant into harmony with its surroundings and give to it a subtle individuality owing to the perfection of their arrangements, structures and forms for the work in hand. The leaves, however, accomplish this work so quietly and economically that most people are scarcely conscious of it”1  Leaves are indeed a great creation of nature and assemble a complex arrangement of synergic collaboration within tissues, which is modified according to their rapidly changing surroundings. On a bigger scope, plants have evolved anatomical structures to take advantage of the place where they live. This is determined mainly by the quality and quantity of sunlight they have, the air and soil temperature, as well as the soil type and associated microorganisms with whom they live. In this context, plants need to adjust in form and function to overcome competition from other plants that may be morphologically better adapted to those site-specific conditions. All these adaptations will influence their physiological performance (e.g., light harvesting, carbon allocation, etc.), and may have a cost in terms of growth and development. These adaptations have led to anatomical characteristics that are species-specific. Indeed, it is possible to find differences within one species, and even within the same individual at different developmental stages. Several studies examining the anatomical variations of leaves have tried to fill out some of the gaps that physiological, genetic and biochemical studies have pointed out (Cormack 1950; Meidner 1953; Friend 1961; Evans et al. 1994; Hanba et al. 2001; Yano                                                  1 Carlton, C.C., 1914. Nature and Development of Plants, 3rd ed., H. Holt and Company, New York. 506 pp.  2 and Terashima 2001; Evans et al. 2009; Scafaro et al. 2011). Studies examining the relationships between structure and function that associate key anatomical and ecophysiological traits in plants are of great importance. This is becoming more frequent, in large part, because through fairly simple leaf structural measurements it is possible to obtain information about the carbon exchange capacity of different plant communities (Niinemets 1999; Slaton and Smith 2002; Tosens et al. 2012). On a global scale (biome-related trends), one of the six most important features of leaves that drives leaf economics is leaf mass area (LMA) (Garnier et al. 1999; Niinemets 1999; Wright et al. 2004b). This trait accounts for the leaf dry-mass investment per unit light-capture area. Species with high LMA tend to have thicker leaves or denser tissues, or both, mainly because they allocate more resources to build structural components that improve physical resistance to abiotic and biotic stress (Muir et al. 2014). In contrast, species with low LMA make a lower investment per unit leaf area, allocating resources to favor rapid growth, and therefore improving competitiveness. Such “choices” set the optimal LMA for each leaf, plant or community (Poorter and Bongers 2006; Muir et al. 2014). Soolanayakanahally (2010) reported an increase in photosynthetic capacity accompanied by an increase in nitrogen content (Na) in northern genotypes of Populus balsamifera L. (balsam poplar). Associated to this was an increase in leaf mass per unit leaf area (LMA), which was larger than the change in Na. The increase in LMA at higher latitudes was assumed to indicate an increased leaf thickness, but might also be explained by a change in density. Changes in density may reflect variation in cell wall thickness, or perhaps something related to the configuration of the tissues in the leaf; for instance, the size and  3 shape of palisade or spongy mesophyll cells, or relative amounts of the two. The present study attempts to resolve some of these matters. 1.1. AgCanBaP germplasm Populus species in general have outstanding advantages over other tree species in terms of their ease of propagation and access to large native collections (Soolanayakanahally 2010). Balsam poplar has an extremely wide distribution across North America (Figure 1), from the Atlantic coast to the Great Lakes and west to Alaska. In fact, it is the northernmost tree species on the continent, extending its range into one of the harshest ecosystems, the arctic (Saarela et al. 2012). The genetic material used in this study originated from geographically distant parent trees in the Agriculture Canada Balsam Poplar (AgCanBaP) collection (Figure 1). The AgCanBaP collection is a range-wide plant collection, and includes germplasm from 62 different locations in Canada and the United States (Soolanayakanahally 2010).   4  Figure 1. The native range (orange) of Populus balsamifera, indicating locations sampled to construct the AgCanBaP collection (green dots). Source: Soolanayakanahally (2010)  1.2. Linking anatomy with physiology  The mesophyll conductance, the ease with which CO2 diffuses from substomatal cavities to sites of carboxylation, is termed gm (Flexas et al. 2008). This conductance is influenced by the pathway length, and the diffusion characteristics present along this pathway, which includes gas-, liquid- and lipid-phase components (Evans et al. 2009; Terashima et al. 2011). Together, the CO2 concentration gradient and gm, directly determine the rate of CO2 diffusion within the leaf (Figure 2, 3a and 3b).  5   Figure 2. Light micrograph of a Populus balsamifera leaf cross-section showing the diffusion of CO2 into leaves. This diffusion (yellow arrow) is driven by differences in CO2 concentration (C, in red) from the airspace outside the stoma (Ca) to the intercellular space (Ci), as controlled by the stomatal conductance (gs). Ca Ci  6 Figure 3. Diffusion pathways into chloroplasts. a) Transmission electron micrograph of palisade cells demonstrating parallel diffusion pathways within cells (indicated with yellow arrows) determined by the exposure of chloroplasts to intercellular air space. b) Transmission electron micrograph of the chloroplast circled in green in the first panel. Parts of the diffusion pathway internal to the leaf reduce the overall conductance, resulting in a drawdown of CO2 concentration from the sub-stomatal space to the outer surface of cell walls (Ci,w) and ultimately the sites of carboxylation within chloroplasts (Cc). The CO2 drawdown from the outer surface of cell walls to chloroplasts (Ci,w-Cc) is determined by the liquid-phase diffusion conductance, gliq (Flexas et al. 2012), one of the components of gm.    ci,w cc a) b)  7 The mesophyll conductance is likely determined by both fixed (e.g., structural anatomy) and variable (e.g., water content, aquaporin and carbonic anhydrase activities, etc.) components that play important roles in the physiological performance of species (Terashima et al. 2006; Flexas et al. 2008; Kaldenhoff 2012). Several studies have tried to estimate the amount of conductance within different parts of the mesophyll, seeking the smallest values of gm, in order to establish the structural barriers and associated consequences on CO2 assimilation. Flexas et al. (2008) noted that cell architecture sets the limit for maximum gm, principally due to the changes that take place in the conductance of the intercellular airspace (gias) and the conductance through cell walls (gw). The gias is mostly determined by mesophyll thickness and porosity (Peguero-Pina et al. 2012; Tosens et al. 2012). Although structural features of the diffusion pathway likely set the upper limit for gm, variable resistances may determine its actual operational value. It has been established that gm is very dynamic, and that it changes within leaves in response to environmental influences as rapidly, or even more rapidly, than stomatal conductance (Flexas et al. 2008). Regarding leaf development, gm and photosynthesis have been shown to increase with leaf photosynthetic capacity coincident with age-dependent modifications in mesophyll cells that increase the exposed chloroplast surface area, and consequently enhance diffusion (Loreto et al. 1994; Hanba et al. 2001; Miyazawa and Terashima 2001). In terms of the drawdown in CO2 concentration from the intercellular air space to the chloroplast (Ci-Cc), it changes from higher Ci-Cc in young leaves to lower Ci-Cc in mature leaves (Loreto et al. 1994; Hanba et al. 2001; Eichelmann et al. 2004).  Likewise,  8 nitrogen content per unit leaf area decreases during leaf development, because the initial N present in young leaves is spread out as the leaf expands (Niinemets et al. 2004; Tosens et al. 2012). 1.3. Leaf mass per unit area (LMA) Leaf mass per unit area is a favored trait to look at in plants because of its ease of measurement and the great variation found between and within species. Unfortunately, interpreting this variation is not always straightforward. LMA is a very important trait as it has been shown to correlate with gm, both positively (Soolanayakanahally et al. 2009; Soolanayakanahally 2010; Tosens et al. 2012) and negatively (reviewed by Flexas et al. 2008). A positive correlation between LMA and gm seems counterintuitive because an increase in leaf thickness should increase the diffusion path length, thereby decreasing gm. On the other hand, an increase in tissue density could increase the cell wall area available for liquid phase diffusion, thereby improving gm. Clearly, variation in LMA needs to be interpreted in terms of variation in leaf thickness and/or leaf density (Niinemets 1999; Evans and Poorter 2001; Wright et al. 2004a). These two factors will be affected at the same time by anatomical tissue traits such as cell size, cell shape, leaf porosity, etc. Leaf density, also referred to as leaf dry matter content, is less commonly measured than LMA but has advantages related to its independency from leaf thickness and more immediate relationship to the allocation of carbon and nutrient resources (Wilson et al. 1999). In a very extensive anatomical study of 26 woody species (Villar et al. 2013), greater leaf thickness in evergreen species was related to higher LMA. Leaves were thicker in  9 evergreen species because of a larger volume of both mesophyll cells and intercellular air space. The mesophyll of evergreen species also had a larger fraction of intercellular air space (fias), when compared to deciduous trees. Nevertheless, within these two groups (evergreen and deciduous), higher LMA was also a result of higher cell density, and consequently, lower fias (Villar et al. 2013). These relationships were validated in a study of evergreen tree species, where thicker leaves had greater LMA and a larger surface area of mesophyll cells exposed to the intercellular air space (Smes/S), but lower fias (Hanba et al. 2014). Despite these known differences, commonalities between species have also been reported. For instance, Slaton and Smith (2002) reported that Deschampsia cespitosa, Oxyria digyna, Hymenoxys grandiflora and Baileya multiradiata had similar values of fias ≈ 0.14, although they occupied different sites (meadow, slope, ridge and desert, respectively) and are taxonomically diverse. 1.4. Nitrogen content (N) Foliar nitrogen content (per unit area and per unit dry mass) tends to be highly correlated with photosynthetic capacity (Niinemets 1999), reflecting the fact that a large fraction of leaf nitrogen is directly involved in photosynthesis. A high biochemical capacity for photosynthesis will tend to increase Ci-Cc by drawing down the CO2 concentration at the sites of carboxylation. Although the drawdown in Cc enhances the rate of diffusion it also limits the potential photosynthetic rate. A concomitant increase in gm would serve to maintain Cc. Hence the expression of changes in gm, or structural features thought to be related to gm, relative to foliar nitrogen content, is of interest. In the context of this study, nitrogen density (Na) and nitrogen density per mesophyll surface area exposed to intercellular air space (Na/Smes/S) are considered.  10 1.5. Connections between traits Slaton and Smith (2002) have emphasized that it is functionally important to consider the joint effects of multiple leaf traits (e.g., fias, leaf thickness, palisade and spongy mesophyll wall areas, etc.), as parts of an interconnected and complex system, rather than just considering them separately. In their work on diverse herbaceous species, these authors sought insight into known connections between the amounts of cell surface area exposed to intercellular space expressed per unit leaf area (Smes/S), and photosynthesis per unit leaf area. They found that leaf thickness accounted for the greatest influences on the Smes/S, fias and the mesophyll cell anatomy and density (Slaton and Smith 2002). However, it was also found that fias, rather than leaf thickness alone, was the best predictor of the chloroplast area available for CO2 absorption during photosynthesis.   Regarding mesophyll cell density (number of cells per unit volume), high density would enhance Smes/S (and the photosynthetic capacity), but it would also reduce the fias and increase the tortuosity of the pathway for CO2 diffusion, thereby decreasing the gias (Niinemets 1999). In contrast, high fias values associated with reduced cell density will reduce the photosynthetic capacity, but increase gias, and potentially, the efficiency of CO2 uptake (Loreto et al. 1994; Garnier et al. 1999). Along the same lines, small cells will provide more surface area per unit volume for CO2 uptake, increasing gw. However, the tortuosity and length of the diffusion pathway through the intercellular air space could decrease the rate of CO2 intake, so there is an implicit trade-off (Slaton and Smith 2002). Past work has found associations between leaf thickness and gm, both positive (Garnier et al. 1999; Niinemets 1999) and negative (Flexas et al. 2008). Nevertheless, the association  11 between Smes/S and photosynthesis is reported to be stronger than the relation between thickness and photosynthesis (Chabot and Chabot 1977; Romero-Aranda et al. 1997). The increment in Smes/S enhances the liquid phase conductance of CO2 into mesophyll cells, facilitating CO2 uptake (Evans 1983). The liquid phase of the CO2 diffusion pathway into chloroplasts includes transfer across the cell wall, passage into the protoplast, diffusion through the cytoplasm, and diffusion across chloroplast membranes. Crucial aspects of this portion of the pathway are the exposed cell wall area, wall thickness, chloroplast positioning, and the exposed chloroplast surface area (Slaton and Smith 2002). Under the premise that gm should be most strongly related to the exposed chloroplast surface area (Sc/S), Evans et al. (1994) conducted a study on transgenic tobacco but found no reduction in CO2 transfer as a result of having more closely packed cells, thicker leaves, nor increased chloroplast thickness. 1.6. Leaf anatomy in poplar There is considerable recent and wide-ranging work in the literature that focuses on poplar species as model systems. Included are many novel works related to leaf anatomy, such as the one carried out by Tosens et al. (2012). These authors examined morphological changes in developing leaves of unstressed European aspen (Populus tremula L.), pursuing anatomical variation linked to the CO2 conductance of mesophyll tissue under different light regimes. In the same study, the authors did a systematic description of the development of the mesophyll, from small, undifferentiated, and tightly packed cells to larger and well-characterized cells, as they differentiated into palisade or spongy mesophyll. As the leaves grew, the fias and the chloroplast surface area were greatly increased, thus improving gm.   12 In the case of balsam poplar, previous work has indicated a positive population level relationship between gm and LMA across the large range of the species (Soolanayakanahally et al. 2009). Northern accessions within the AgCanBaP collection  tend to have higher gm associated with higher LMA, higher rates of photosynthesis, less carbon isotope discrimination (higher water-use efficiency) and higher chlorophyll content readings (Soolanayakanahally et al. 2009). Families generated by crossing northern accessions with southern accessions present considerable variation in both LMA and gm, which are positively associated (Ryan, in preparation). In contrast, in black cottonwood (Populus trichocarpa Torr. and Gray), northern accessions also had higher photosynthesis (and perhaps greater isotope discrimination) but LMA and chlorophyll content were better correlated with environmental dryness (McKown et al. 2014a). Leaf thickness per se was not measured in either of these studies, but in both cases it was presumed to account for most of the variation in LMA.   1.7. Research hypothesis and objectives The present study used Populus balsamifera as a system to explore relationships between LMA and leaf anatomy that may bear on variation in gm. The hypotheses driving the research, and associated objectives that were pursued, are as follows:  h1: LMA varies directly with thickness of the leaf, and is a suitable proxy measure of such o1: Establish the relationship between leaf mass area (LMA) and leaf anatomy  h2: Leaves with higher LMA will have more cell wall area for CO2 diffusion to chloroplasts, or lower nitrogen per unit exposed cell wall area  13 o2: Quantify cell wall area, chloroplast area and leaf nitrogen in different genotypes and relate these to LMA  h3: gm varies as a function of leaf microanatomy and/or ultrastructure o3: Analyze the relationship between leaf thickness, LMA and other structural limitations that might influence gm.  14 CHAPTER 2.  Material and methods  2.1. Plant material The genetic material used in this study originated from geographically distant parent trees belonging to populations sampled to create the AgCanBaP collection (Table 1; note that the parental trees were not subsequently propagated as part of the AgCanBaP collection). A cross between a female tree (K4) of subarctic origin (Kuujjuaq QC) and a polymix of pollen from three males (C) of prairie origin (Carnduff SK) resulted in a family with high variation in LMA (Ryan, in preparation). When the breeding was conducted, pollen from the three fathers was mixed together (also known as a polymix) in order to have enough to pollinate every mother. Therefore, some of the members of the K4×C family are full siblings and some are half-sibs, but the exact parentage of each genotype remains unknown. Table 1. Geographic coordinates of the parental material  Population Latitude (oN) Longitude (oW) Elevation (m) Carnduff 49.18 101.83 558 Kuujjuaq 58.02 68.65 17  Seventeen K4×C progeny were used to grow plants in a glasshouse under an extended-day-regime for 12 weeks from May to July 2012. Dormant whips of 20 cm, with two buds per whip, were forced to root in 1-gallon pots filled with a mixture of peat moss (70%) and Perlite (30%) (West-Creek Farms, Langley, Canada). There were five  15 replicates per genotype, for a total of 85 pots (Figure 4). Pots were randomized weekly to minimize positional effects.   Figure 4. Development of the plants at the greenhouse over the first 8 weeks of growth (a to h, respectively).  The rooted cuttings were grown with natural light supplemented by 400-W high-pressure sodium (HPS) lamps to provide a 21-hour-photoperiod and a minimum photosynthetic photon flux density (PPFD) of 677 µmol m-2 s-1. Maximum day and night temperatures  16 were maintained at approximately 24 and 20°C, respectively. After flushing, the plants were well watered and fertilized using Plant-Prod 15-5-15 (N-P-K) plus micronutrients (boron, copper, iron, manganese, molybdenum and zinc) at a rate of 100 ppm N. Sixty-nine cuttings grew successfully and were kept in good condition. After one month some flies were found on the plants and, for this reason, a Koppert Biological Systems (Koppert Canada Ltd, Scarborough ON) EN-STRIP Encarsia formosa parasitic wasp biological control agent was applied. In addition, during the last week of the experiment some replicates died from a brief, inadvertent drought stress. The youngest fully mature leaf of each plant was harvested after three months growth. They were severed in the morning between 10:00 and 12:00 h and stored for 30 minutes in a labeled bag with a wet paper towel inside to prevent water loss. The collected leaf samples were subsampled for stomatal impressions and for determination of dry weight and elemental analysis. They were processed for microscopy analysis after a day of storage in a cold room at 4°C.  2.2. Developmental stage of the plants As mentioned above, each harvested leaf was selected to obtain the youngest mature leaf from each plant. However, the different genotypes grew at different rates and leaves reached visual maturity at different levels of insertion from the shoot tip. Consequently, to obtain further information, Leaf Plastochron Index (LPI), which is meant to provide a rough idea of leaf developmental (Erickson and Michelini 1957), was registered for each harvested leaf according to Erickson and Michelini (1957):  17  𝐿𝑃𝐼 = 𝑛 + log 𝐿𝑛−log10(log𝐿𝑛−log𝐿𝑛+1)  [1] where, n is the serial number of that leaf which just exceeds 15 mm in length, and log Ln is the logarithm of length in mm of leaf n. A reference length of 15 mm was chosen for LPI calculation (LPI=0), since leaves of this length grow exponentially and are also long enough to be measured without damaging the growing shoot.  2.3. Leaf mass per unit area, carbon and nitrogen content (LMA, C%, N%) Two leaf tissue discs (total area: 61 mm2) were taken using a hand-held punch. Discs were oven dried at 50°C for 48 hours and weighed to determine LMA (mg mm-2). Between 2-2.5 mg of dried tissue per replicate (2-5 per each genotype), was analyzed for carbon (%C) and nitrogen (%N) content at the Stable Isotope Facility in the Department of Forest and Conservation Sciences, Faculty of Forestry, UBC. Homogenization of samples, when necessary, was performed manually with a mortar and pestle.  2.4. Chlorophyll concentration index (CCI) An indication of the foliar chlorophyll content was obtained for each harvested leaf prior to collection using a CCM-200 plus Chlorophyll Content Meter (Opti-Sciences, Inc., USA). This instrument reports chlorophyll content as a unit-less index based on the leaf transmittance of red (650 nm) and infrared (940 nm) wavelengths. Chlorophyll measurements were taken in the middle section of the leaves.    18 2.5. Stomatal density (stoD) and length of the stomatal aperture (stoLA) Balsam poplar only has stomata on the abaxial side of the leaves (hypostomatous); this fact was confirmed with the light microscopy images. stoD was obtained by making impressions with clear nail polish applied to one side of the mid-rib near the middle of the abaxial surface. The counting was done through Image J (Rasband 2014) in at least 10 (up to 25) fields of view per genotype, and recorded at 40× magnification (McKown et al. 2014b). Length of the stomatal aperture (stoLA), was also measured for the same images in at least 100 stomata, up to a maximum of 250. This length is found between the junctions of the guard cells at each end of the stoma. Stomatal impressions were completed for 13 genotypes only (K4C1, K4C2, K4C3, K4C4 were not sampled) because of tissue limitations and the loss of some peels.  2.6. Microscopy For microscopy analysis, one piece of tissue of about 2x1 mm was cut from the mid-portion of each harvested leaf, making sure to avoid the midrib and larger veins. The excised tissues were fixed in a solution of 2.5% glutaraldehyde and 0.05M sodium cacodylate (pH 6.9), and placed in the fridge until they were ready for embedding.  Samples were fixed with 2.5% glutaraldehyde in 0.05 M sodium cacodylate for two cycles of 2 min on, 2 min off and 2 min on microwave power level 1 (PELCO3240 Microwave Load Cooler at 22.9°C). Samples were then post-fixed with 1% OsO4 in 0.05 M sodium cacodylate for 1 hour. Samples were subsequently placed under vacuum for 5 min to remove any remaining air in the leaves. Once vented, the samples were rinsed three times with distilled water.  19 Samples were then microwave dehydrated for 1 min per step in a graded series of ethanol at 30, 50, 70, 90, 95, and 3×100%. This was followed by infiltration with Spurr's resin (Spurr 1969) at 25, 50, 75, and 3×100% in the microwave for 3 min at power level 3, followed by 30 min on a rotator.  Transverse sections approximately 0.5 μm and 50-70 nm thick (or when a grey ribbon was achieved) were obtained using a microtome and a diamond knife for light and electron microscopy, respectively.  2.6.1. Microanatomy by light microscopy Tissue-level measurements were completed on thick sections of 0.5 μm stained with Toluidine Blue O prior to imaging. Six technical replicates were analyzed for each trait using Image J.  2.6.1.1 Leaf and tissue thicknesses Total leaf (tleaf), mesophyll (total thickness excluding adaxial and abaxial epidermal layers), palisade (tpal), and spongy mesophyll (tspo) thicknesses were measured from images of 20× magnification, and expressed in mm. 2.6.1.2. Cross-sectional areas Palisade, upper and lower spongy mesophyll, and abaxial and adaxial epidermal cross-sectional surface areas (Ap, Aupp, Alow, Aab and Aad, respectively) were measured on cropped images of 0.134 mm × 0.261 mm at 40× magnification. The areas were expressed in mm2 mm-1 (i.e., cross-sectional area per mm of leaf width). The absence of  20 chloroplasts in the lower spongy mesophyll was used to determine the distinction between upper and lower spongy mesophyll. 2.6.1.3. Fraction of mesophyll occupied by intercellular air space The fraction of leaf volume consisting of intercellular air space (fias) was calculated according to Evans et al. (1994):   𝑓ias = 1 −∑𝑆s𝑡mes∗𝑊 [2] where, ∑Ss is the cross-sectional area of the mesophyll cells, tmes is the mesophyll thickness between the two epidermal layers, and W is the width of the section being measured. 2.6.1.4. Ratios of the exposed and total surface areas of mesophyll cells per unit area of leaf surface To calculate the mesophyll cell wall surface area exposed to intercellular air space (Smes/S) and the total surface area of the mesophyll (Pmes/S), cell perimeters and adjoining lengths were first measured on cropped images (0.134 mm × 0.261 mm at 40× magnification). Smes/S was calculated according to (Evans et al. 1994; Parkhurst 1994; Evans and Caemmerer 1996; Hanba et al. 2001; Scafaro et al. 2011; Tosens et al. 2012; Hanba et al. 2014), as follows:  𝑆𝑚𝑒𝑠/𝑆 =𝐿𝑚𝑒𝑠𝑊∗ γ [3]  21 where, Lmes is the total length of mesophyll cell walls facing the intercellular air space (mm), W is the width of the section analyzed (mm), and γ is the curvature correction factor. To calculate Pmes/S, total perimeter of the cells was used instead of Lmes. In Equation 3, γ accounts for the fact that in general the surface (e.g., of a palisade cell) is not perpendicular to the section. The value of γ depends on the shape of the cells under consideration, as a weighted average of its axis for each of palisade, upper spongy, and lower spongy mesophyll tissues (Thain 1983). The value can be obtained by assuming that the cells are spheroids (either oblate or prolate spheroid) having different width to height ratios (Thain 1983). The average γ for palisade cells was 1.47, and for upper and lower spongy mesophyll cells it was 1.35. 2.6.1.5. Chloroplasts in mesophyll tissues Numbers of visible chloroplasts in palisade (Chlo_P) and spongy mesophyll (Chlo_US) cells were counted in 10 cells of each type on cropped images at 40× magnification.   2.6.2. Ultrastructure by transmission electron microscopy (TEM) TEM was used for subcellular measurements, such as mesophyll cell wall thickness (tcw), chloroplast surface area exposed to intercellular airspace per unit of leaf area (Sc/S), and thylakoid grana thickness. Ultra-thin sections for TEM were contrasted with both uranyl acetate (30’) and Pb citrate (15’), and mounted in Formvar-carbon grids prior to imaging. Images were examined in a Hitachi H7600 with a 120kV tungsten filament. Multiple electron micrographs were measured for each leaf, sampling both palisade and spongy mesophyll tissues.  22 2.6.2.1. Cell wall thickness (tcw) Images with 40,000× magnification were used for tcw estimations. Ten measurements per each of the six technical replicates were done throughout all the cell wall (μm) (Tosens et al. 2012). 2.6.2.2. Thylakoid grana thickness (tgra) The thickness of the thylakoid grana was also determined following the same procedure used to measure cell wall thickness. 2.6.2.3. Chloroplast surface area exposed to intercellular air space (Sc/S) Sc/S was calculated from electron micrographs as shown in (Evans et al. 1994),  𝑆𝑐/𝑆 =𝐿𝑐′𝐿𝑚𝑒𝑠′∗𝑆𝑚𝑒𝑠𝑆 [4] where, Lc' is the total length of the chloroplasts exposed to the intercellular air space on the cropped images (18.47 μm × 21.42 μm at 8000× magnification); Lmes' is the total length of mesophyll cell walls facing the intercellular air space; and Smes/S is the mesophyll surface area exposed to intercellular air space. 2.7. Statistical analyses One-way nested analyses of variance (ANOVA) were run using SAS version 9.4 to test for effects of different genotypes. Extreme observations (outliers) were removed in order to meet the normality and uniform variance assumptions for Aupp, Ap, Aab, Smes/S_T, fias, Pmes/S_US, Pmes/S_LS Smes/S_P, Smes/S_US, Smes/S_LS, stoD and stoLA. Although transformation was applied for tgra and tcw, both assumptions were not met for those traits.  23 ANOVAs were considered significant at p<0.05. Multiple comparisons were done following Bonferroni’s procedure following Kozak et al. (2008). Tables of clonal mean values, standard deviations and results of multiple comparisons are presented when genotypes differed significantly. Data for traits where no differences were found are presented graphically. Pearson’s product-moment correlation coefficients between selected traits were calculated using SigmaPlot 12.3. This software was also used to regress LMA on Pmes/S and for multiple linear regression of LMA on tleaf and fias.   24 CHAPTER 3. Results   3.1. Developmental stage of the plants 3.1.1. Leaf plastochron index (LPI) The plant material LPI ranged from values of six up to a maximum of 11 (Table 2), with an overall mean of 9. The high standard deviation in K4C13 shown in Table 2 may be partly due to the low sampling of this particular genotype (many of them could not survive the greenhouse conditions, and died from drought). The other fluctuations came from the variation within clones. Genotype LPI ranged from a low of 6.515 in K4C8, to a high of 11.667 in K4C20 (Table 2). Table 2. Analysis of variance of the leaf plastochron index (n=2 to 5) C LPI ANOVA SD K4C8 6.515 a 2.075 K4C3 7.385 ab 0.768 K4C19 7.408 abc 1.126 K4C17 7.522 abcd 3.103 K4C14 7.712 abcde 1.498 K4C11 7.906 abcdef 2.334 K4C1 8.061 abcdefg 0.660 K4C4 8.625 abcdefgh 1.054 K4C7 8.714 abcdefghi 2.398 K4C15 9.148 abcdefghij 1.215 K4C18 9.246 abcdefghijk 1.577 K4C12 9.303 abcdefghijkl 1.215 K4C9 10.050 abcdefghijklm 1.612 K4C6 10.137 abcdefghijklmn 0.353 K4C2 10.965 abcdefghijklmno 2.643 K4C13 11.443 abcdefghijklmnop 3.138 K4C20 11.667 bcdefghijklmnop 1.692 C: clones, SD: standard deviation, LPI: leaf plastochron index.   25 3.2. Whole tissue properties 3.2.1. Leaf mass area (LMA) Most of the genotypes had similar LMA values, and those that did not (K4C8, K4C9) also had very high standard deviation (Figure 5).   Figure 5. Leaf mass per unit area, by clone. Error bars are ±SD. n = 2 (hence no error bars) for K4C-1, 6 and 13; n = 3 for K4C12; n = 4 for K4C-2, 8, 9, 11, 17, 20 and n= 5 for the rest.  3.2.2. Carbon-Nitrogen ratio (C:N) and Nitrogen density (Na) Carbon-Nitrogen ratio was fairly similar among all genotypes (Figure 6), except for K4C1 and K4C19, which had the largest values. Nitrogen content per unit leaf area was also fairly constant among clones (Figure 7), except for K4C8 and K4C9, which had the lowest values, but also the highest standard deviations.  00.010.020.030.040.050.06K4C1 K4C2 K4C3 K4C4 K4C6 K4C7 K4C8 K4C9 K4C11 K4C12 K4C13 K4C14 K4C15 K4C17 K4C18 K4C19 K4C20LMA (mg mm-2)Clones 26  Figure 6. Carbon-Nitrogen ratio, by clone. Error bars are ±SD. n = 2 for K4C-1, 6 and 13 (no error bars); n = 3 for K4C12; n = 4 for K4C-2, 8, 9, 11, 17, 20 and n= 5 for the rest.   Figure 7. Nitrogen content per unit area, by clone. Error bars are ±SD. n = 2 for K4C-1, 6 and 13 (no error bars); n = 3 for K4C12; n = 4 for K4C-2, 8, 9, 11, 17, 20 and n= 5 for the rest.   02468101214K4C1 K4C2 K4C3 K4C4 K4C6 K4C7 K4C8 K4C9 K4C11 K4C12 K4C13 K4C14 K4C15 K4C17 K4C18 K4C19 K4C20C:N (mg mg-1 )Clones00.511.522.5K4C1 K4C2 K4C3 K4C4 K4C6 K4C7 K4C8 K4C9 K4C11 K4C12 K4C13 K4C14 K4C15 K4C17 K4C18 K4C19 K4C20Na(g m-2)Clones 27 3.2.3. Chlorophyll concentration index (CCI) The index of leaf chlorophyll concentration showed a lot of variation among genotypes, but K4C13 stood out (in part because of its relatively high LPI). Several significant differences between clones were found on this trait (Table 3). The CCI varied from a low of 24.02 in K4C19 to a high of 53.23 in K4C13, covering a much greater range than C:N ratio or Na. Table 3. Analysis of variance of the chlorophyll concentration (n=2 to 5) C CCI ANOVA SD K4C19 24.020 a 2.069 K4C8 24.102 ab 8.316 K4C1 29.481 abc 4.438 K4C15 30.700 abcd 4.697 K4C9 31.835 abcde 8.612 K4C3 32.047 abcdef 4.881 K4C6 32.242 abcdefg 3.964 K4C14 32.267 abcdefgh 2.448 K4C11 32.473 abcdefghi 8.282 K4C7 32.947 abcdefghij 5.537 K4C17 34.227 bcdefghijk 7.683 K4C20 34.643 bcdefghijkl 5.819 K4C18 34.880 cdefghijklm 12.097 K4C12 36.048 cdefghijklmn 6.750 K4C4 37.240 cdefghijklmno 3.717 K4C2 41.302 cefghijklmnop 8.866 K4C13 53.229 p 16.138  C: clones, SD: standard deviation, CCI: genotypic means of the chlorophyll concentration.   28 3.3. Microanatomy The light micrographs allowed measurement of different tissues in the leaves, such as the epidermal layers (adaxial and abaxial), the palisade tissue, and lastly, the upper and lower spongy mesophyll tissues (Figure 8).   Figure 8. Light micrographs of K4C9 and K4C13 indicating thickness measurements within the leaf and morphological differences between the two clones. abaxial adaxial palisade upper spongy lower spongy leaf thickness  29 3.3.1. Leaf thickness (tleaf) and its components Significant clonal differences in thickness of the leaf, as well as the palisade and spongy mesophyll, were found (Table 4). The depth of the spongy mesophyll was more variable and typically exceeded the depth of the palisade. Thicker leaves tended to have both a thicker palisade and a thicker spongy mesophyll. 3.3.2. Cross-sectional areas (Ap, Aupp, Alow, Aab and Aad) The tissues present in the mesophyll were easily defined (as previously shown in Figure 8), with one layer of adaxial and one layer of abaxial epidermal cells, two layers of palisade cells, and two less regularly distributed zones of spongy mesophyll (upper and lower). The palisade region presented the biggest cross-sectional area, excluding intercellular air space, within the total mesophyll tissue. The spongy mesophyll was much less packed and condensed, with a larger fraction of air space (especially in the lower spongy mesophyll) when compared with palisade tissue. Several significant differences were found between genotypes in all tissue layers of the leaf (Table 5 and Table 6). Genotypes with a large palisade cross-sectional area (Ap) tended also to have a large upper spongy mesophyll cross-sectional area (Aupp), but not necessarily a large lower spongy mesophyll cross-sectional area (Alow). 3.3.3 Fraction of mesophyll occupied by intercellular air space (fias) Several differences among genotypes were found in fias, which ranged from a low of 0.4730 in K4C19, to a high of 0.6130 in K4C9 (Table 7). The fias would of course also vary with each tissue type, in accordance with their variation in cross-sectional area and thickness, as presented above.  30 Table 4. Analysis of variance for leaf, palisade and spongy mesophyll thickness (n=2) C tleaf  ANOVA SD C tpal ANOVA SD C tspo ANOVA SD K4C8 0.179 a 0.026 K4C19 0.062 a 0.002 K4C3 0.063 a 0.021 K4C3 0.184 ab 0.032 K4C9 0.066 ab 0.006 K4C8 0.064 ab 0.017 K4C19 0.190 b 0.011 K4C15 0.066 abc 0.007 K4C4 0.072 abch 0.009 K4C14 0.198 c 0.013 K4C8 0.066 abcd 0.009 K4C9 0.073 abcd 0.008 K4C4 0.202 cd 0.008 K4C6 0.067 abcde 0.003 K4C11 0.076 abcde 0.012 K4C17 0.203 cde 0.023 K4C3 0.068 bcdef 0.004 K4C14 0.077 abcdef 0.011 K4C15 0.204 cdef 0.012 K4C17 0.068 bcdefg 0.006 K4C20 0.078 abcdefg 0.010 K4C9 0.204 cdefg 0.016 K4C1 0.069 bcdefgh 0.004 K4C17 0.079 bcdefgh 0.013 K4C11 0.206 defgh 0.037 K4C12 0.069 bcdefghi 0.003 K4C19 0.079 bcdefghi 0.005 K4C12 0.207 defgh 0.003 K4C14 0.070 bcdefghij 0.002 K4C7 0.083 cdefghij 0.013 K4C2 0.214 i 0.005 K4C2 0.073 fghijk 0.006 K4C2 0.083 cdefghijk 0.006 K4C20 0.216 ij 0.003 K4C18 0.076 kl 0.003 K4C15 0.084 cdefghijkl 0.007 K4C6 0.221 jk 0.013 K4C11 0.076 klm 0.010 K4C6 0.085 cdefghijklm 0.007 K4C18 0.227 k 0.018 K4C7 0.076 klmn 0.012 K4C12 0.088 defghijklmn 0.003 K4C1 0.236 l 0.015 K4C4 0.077 klmno 0.004 K4C18 0.090 efghijklmno 0.010 K4C7 0.236 l 0.025 K4C20 0.081 lmnop 0.005 K4C1 0.104 op 0.008 K4C13 0.254   0.006 K4C13 0.081 op 0.010 K4C13 0.110 p 0.011 C: clones, SD: standard deviation, tleaf: genotypic mean of leaf thickness (mm), tpal: genotypic mean of palisade thickness (mm), tspo: genotypic mean of spongy mesophyll thickness (mm).  31 Table 5. Analysis of variance of cross-sectional cell area within mesophyll tissues, per mm of leaf surface (n=2) C Ap ANOVA SD C Aupp ANOVA SD C Alow ANOVA SD K4C9 0.047 a  0.004 K4C8 0.017 a  0.006 K4C3 0.007 a 0.002 K4C8 0.047 ab 0.013 K4C3 0.019 ab 0.005 K4C4 0.008 ab 0.003 K4C19 0.048 abc 0.001 K4C9 0.020 abc 0.003 K4C9 0.009 abc 0.002 K4C6 0.050 abcd 0.003 K4C4 0.021 abcd 0.003 K4C8 0.009 abcd 0.003 K4C17 0.050 abcde 0.006 K4C7 0.023 abcde 0.007 K4C7 0.009 abcde 0.001 K4C14 0.051 abcdef 0.003 K4C14 0.023 abcdef 0.003 K4C11 0.010 abcdef 0.002 K4C3 0.051 abcdefg 0.012 K4C2 0.024 abcdefg 0.005 K4C18 0.010 abcdefg 0.004 K4C15 0.052 abcdefgh 0.001 K4C6 0.024 abcdefgh 0.005 K4C14 0.010 abcdefgh 0.003 K4C12 0.052 abcdefghi 0.003 K4C18 0.026 abcdefghi 0.002 K4C17 0.010 abcdefghi 0.004 K4C1 0.053 abcdefghij 0.004 K4C20 0.026 abcdefghij 0.004 K4C12 0.011 abcdefghij 0.002 K4C11 0.055 abcdefghijk 0.008 K4C17 0.026 abcdefghijk 0.009 K4C6 0.011 abcdefghijk 0.002 K4C4 0.056 abcdefghijkl 0.003 K4C11 0.027 abcdefghijkl 0.003 K4C2 0.013 abcdefghijkl 0.002 K4C20 0.058 abcdefghijklm 0.006 K4C12 0.027 abcdefghijklm 0.013 K4C15 0.014 bcdefghijklm 0.005 K4C18 0.059 abcdefghijklmn 0.003 K4C19 0.030 cdefghijklmn 0.005 K4C20 0.016 fghijklmn 0.006 K4C2 0.059 abcdefghijklmno 0.004 K4C1 0.030 defghijklmno 0.004 K4C19 0.016 ghijklmno 0.005 K4C7 0.060 abcdefghijklmnop 0.005 K4C15 0.031 defghijklmnop 0.005 K4C1 0.016 ghijklmnop 0.005 K4C13 0.067 klmnop 0.006 K4C13 0.034 efghijklmnop 0.006 K4C13 0.018 lmnop 0.002 C: clones, SD: standard deviation, Ap: genotypic means of the palisade surface area (mm2 mm-1), Aupp: genotypic means of the upper spongy surface area (mm2 mm-1), Alow: genotypic means of the lower spongy surface area (mm2 mm-1).      32 Table 6. Analysis of variance of cross-sectional area of adaxial and abaxial epidermal layers, per mm of leaf surface (n=2).  C Aad ANOVA SD C Aab ANOVA  SD K4C12 0.007 a 0.001 K4C12 0.005 a  0.000 K4C13 0.008 ab 0.001 K4C6 0.006 ab 0.002 K4C3 0.008 abc 0.001 K4C9 0.006 abc 0.001 K4C2 0.008 abcd 0.003 K4C8 0.007 abcd 0.002 K4C17 0.009 abcde 0.001 K4C19 0.007 abcde 0.001 K4C14 0.009 abcdef 0.001 K4C1 0.007 abcdef 0.001 K4C4 0.010 abcdefg 0.001 K4C18 0.007 abcdefg 0.001 K4C1 0.010 bcdefghi 0.000 K4C14 0.008 abcdefgh 0.001 K4C7 0.010 bcdefgi 0.002 K4C7 0.008 abcdefghi 0.002 K4C8 0.010 bcdefghij 0.001 K4C4 0.008 abcdefghij 0.001 K4C15 0.010 defghijk 0.001 K4C11 0.008 abcdefghijk 0.001 K4C6 0.010 defghijkl 0.001 K4C13 0.008 abcdefghijkl 0.002 K4C11 0.010 defghijklm 0.001 K4C15 0.008 abcdefghijklm 0.001 K4C19 0.010 defghijklmn 0.002 K4C3 0.009 abcdefghijklmn 0.002 K4C9 0.010 efghijklmno 0.001 K4C17 0.009 bcdefghijklmno 0.002 K4C18 0.011 efghijklmnop 0.001 K4C2 0.009 bcdefghijklmnop  0.002 K4C20 0.012 hijklmnop 0.001 K4C20 0.009 bcdefghijklmnop  0.002 C: clones, SD: standard deviation, Aad: genotypic means of the adaxial surface area (mm2 mm-1), Aab: genotypic means of the abaxial surface area (mm2 mm-1).    33 Table 7. Analysis of variance of fias values (n=2) C fias ANOVA SD K4C19 0.473 a 0.074 K4C15 0.504 ab 0.036 K4C20 0.511 abc 0.015 K4C13 0.512 abcd 0.011 K4C2 0.530 abcde 0.025 K4C11 0.531 abcdef 0.018 K4C12 0.542 abcdefg 0.018 K4C14 0.555 bcdefgh 0.037 K4C17 0.555 bcdefghi 0.034 K4C18 0.561 bcdefghij 0.041 K4C1 0.561 bcdefghijk 0.024 K4C4 0.561 bcdefghijkl 0.028 K4C8 0.575 bcdefghijklm 0.065 K4C3 0.578 bcdefghijklmn 0.040 K4C7 0.592 cdefghijklmno 0.035 K4C6 0.596 efghijklmnop 0.009 K4C9 0.610 efghijklmnop 0.048 C: clones, SD: standard deviation, fias: genotypic means of the fraction of mesophyll occupied by intercellular air space.  3.3.4. Surface area of mesophyll cells (Pmes/S) and mesophyll surface area exposed to intercellular air space per unit leaf area (Smes/S) The total, three-dimensional cell surface area of the mesophyll (Pmes/S) and its component tissues was calculated from the measured cell wall perimeters. Smes/S was calculated similarly but corrected for that portion where walls of adjacent cells touch. Consequently, Pmes/S exceeded Smes/S, and more so in the palisade than in the spongy mesophyll. When comparing Pmes/S and Smes/S among genotypes, it is important to bear in mind that these traits are determined from a large set of anatomical characteristics, and influenced by the combination of leaf thickness, cell density, cell shape and cell size, and for each tissue layer separately. Several differences among genotypes were found in total (summed across the entire mesophyll) Pmes/S and Smes/S (Table 8) and at the different tissue levels  34 (Tables 9 and 10). Unsurprisingly, differences in total Pmes/S and Smes/S, and their components, tend to parallel each other across genotypes.  Table 8. Analysis of variance of Pmes/S_T and Smes/S_T values (n=2) C Pmes/S_T ANOVA SD C Smes/S_T ANOVA SD K4C8 33.309 a 9.872 K4C8 24.3457 a 6.653 K4C9 34.665 ab 3.934 K4C3 26.8199 ab 4.993 K4C14 38.583 abc 6.541 K4C9 27.3691 abc 3.924 K4C3 40.612 abcd 6.990 K4C17 29.0226 abcd 3.554 K4C17 42.330 abcde 5.887 K4C14 29.0855 abcde 6.485 K4C18 42.966 abcdef 4.434 K4C19 31.2077 abcdef 1.364 K4C11 43.097 abcdefg 6.211 K4C7 31.2992 abcdefg 5.091 K4C7 43.802 abcdefgh 5.175 K4C6 32.4498 abcdefgh 2.967 K4C15 44.137 abcdefghi 4.344 K4C18 32.8566 abcdefghi 2.653 K4C20 44.194 abcdefghij 3.664 K4C1 32.8912 abcdefghij 3.676 K4C4 44.349 abcdefghijk 3.341 K4C12 33.1522 abcdefghijk 3.114 K4C6 45.107 abcdefghijkl 8.413 K4C4 33.1547 abcdefghijkl 2.401 K4C12 45.900 bcdefghijklm 7.785 K4C11 33.4931 bcdefghijklm 4.510 K4C19 46.007 bcdefghijklmn 2.702 K4C15 33.5219 bcdefghijklmn 4.275 K4C1 48.825 cdefghijklmno 4.451 K4C20 34.2044 bcdefghijklmno 3.383 K4C13 52.496 defghijklmnop 4.227 K4C2 34.9392 bcdefghijklmnop 5.316 K4C2 57.074 lmnop 5.021 K4C13 36.3912 defghijklmnop 2.639 C: clones, SD: standard deviation, Pmes/S_T: genotypic means of the total surface area of mesophyll cells calculated from total perimeter of cell walls in mesophyll tissue (mm2 mm-2), Smes/S_T: genotypic means of the total surface area of mesophyll cells exposed to airspace per unit of leaf area in mesophyll tissue (mm2 mm-2).   35 Table 9. Analysis of variance of Pmes/S_P, Pmes/S_US and Pmes/S _LS values (n=2) C Pmes/S_P  ANOVA SD C Pmes/S_US  ANOVA SD C Pmes /S_LS  ANOVA SD K4C8 20.211 a 6.813 K4C8 8.4109 a 2.482 K4C9 4.0392 a 0.673 K4C9 20.683 ab 1.381 K4C6 9.8506 ab 1.611 K4C11 4.4785 ab 0.646 K4C14 24.143 abc 3.234 K4C14 9.8520 abc 2.651 K4C14 4.5882 abc 1.460 K4C15 24.335 abcd 2.105 K4C9 9.9433 abcd 3.768 K4C3 4.6588 abcd 1.887 K4C17 24.371 abcde 3.088 K4C3 10.5458 abcde 1.815 K4C8 4.6874 abcde 1.212 K4C1 25.005 abcdef 1.784 K4C18 11.2837 abcdef 3.643 K4C17 4.9327 abcdef 1.704 K4C3 25.408 abcdefg 5.383 K4C20 11.4275 abcdefg 3.658 K4C18 5.0852 abcdefg 1.756 K4C11 25.795 abcdefgh 4.010 K4C7 11.6006 abcdefgh 3.961 K4C4 5.3729 abcdefgh 2.061 K4C19 26.263 abcdefghi 2.012 K4C4 12.2291 abcdefghi 1.992 K4C20 5.3866 abcdefghi 1.556 K4C6 26.323 abcdefghij 0.925 K4C11 12.8234 abcdefghij 4.019 K4C12 5.7828 abcdefghij 1.118 K4C7 26.508 abcdefghijk 2.378 K4C17 13.0263 abcdefghijk 3.879 K4C15 6.0917 abcdefghijk 2.845 K4C18 26.598 abcdefghijkl 1.276 K4C12 13.1326 abcdefghijkl 3.311 K4C7 6.1351 abcdefghijkl 3.009 K4C4 26.747 abcdefghijklm 1.256 K4C19 13.2338 abcdefghijklm 2.638 K4C6 6.2397 abcdefghijklm 3.482 K4C12 26.984 bcdefghijklmn 4.128 K4C15 13.7096 abcdefghijklmn 2.148 K4C19 6.5108 abcdefghijklmn 1.703 K4C20 27.380 cdefghijklmno 2.257 K4C1 14.3117 abcdefghijklmno 4.479 K4C2 7.3731 abcdefghijklmno 2.679 K4C13 29.902 cdefghijklmnop 1.737 K4C13 14.6465 abcdefghijklmnop 3.899 K4C13 7.9475 abcdefghijklmnop 3.046 K4C2 30.882 defghijklmnop 1.889 K4C2 16.7976 bcdefghijklmnop 1.594 K4C1 9.5084 efghijklmnop 4.239 C: clones, SD: standard deviation, Pmes/S genotypic means of the surface area of mesophyll tissues calculated from total perimeter of cell walls (P: palisade, US: upper spongy mesophyll and LS: lower spongy mesophyll, all in mm2 mm-2).  36 Table 10. Analysis of variance of Smes/S_P, Smes/S_US and Smes/S_LS values (n=2) C Smes/S_P  ANOVA SD   C Smes/S_US  ANOVA SD   C Smes/S_LS  ANOVA SD K4C8 13.224 a  3.345   K4C8 6.765 a 2.619   K4C9 3.857 a 0.640 K4C2 14.053 ab 3.418   K4C3 7.657 ab 0.879   K4C3 4.246 ab 1.693 K4C1 14.557 abc 0.886   K4C14 8.106 abc 1.915   K4C8 4.356 abc 1.286 K4C3 14.918 abcd 4.569   K4C9 8.358 abcd 3.556   K4C11 4.363 abcd 0.618 K4C9 15.154 abcde 1.066   K4C19 8.924 abcde 1.006   K4C14 4.482 abcde 1.447 K4C15 15.155 abcdef 0.487   K4C17 8.969 abcdef 3.567   K4C17 4.621 abcdef 1.543 K4C17 15.433 abcdefg 1.987   K4C18 9.329 abcdefg 2.865   K4C18 4.923 abcdefg 1.732 K4C7 15.965 abcdefgh 2.967   K4C20 9.383 abcdefgh 3.408   K4C4 4.995 abcdefgh 1.834 K4C14 16.498 abcdefghi 3.999   K4C4 9.703 abcdefghi 1.799   K4C20 5.154 abcdefghi 1.477 K4C19 16.824 abcdefghij 1.688   K4C1 9.731 abcdefghij 3.136   K4C19 5.459 abcdefghij 0.753 K4C12 17.297 abcdefghijk 1.295   K4C7 9.831 abcdefghijk 3.564   K4C7 5.503 abcdefghijk 3.032 K4C13 17.936 abcdefghijkl 2.693   K4C12 10.285 abcdefghijkl 2.195   K4C12 5.570 abcdefghijkl 1.079 K4C4 18.457 abcdefghijklm 0.946   K4C11 10.287 abcdefghijklm 3.308   K4C6 5.904 abcdefghijklm 3.219 K4C18 18.604 bcdefghijklmn 1.159   K4C6 10.728 abcdefghijklmn 5.602   K4C2 6.227 abcdefghijklmn 2.094 K4C6 18.697 bcdefghijklmno 1.619   K4C13 11.119 abcdefghijklmno 2.392   K4C15 7.234 abcdefghijklmno 4.293 K4C11 18.843 bcdefghijklmnop 2.916   K4C15 11.133 abcdefghijklmnop 2.351   K4C13 7.337 abcdefghijklmnop 2.716 K4C20 19.667 cdefghijklmnop 1.596   K4C2 14.659 cdefghijklmnop 4.450   K4C1 8.603 bcdefghijklmnop 3.579  C: clones, SD: standard deviation, Smes/S: genotypic means of the surface area of mesophyll cells exposed to airspace per unit of leaf area in mesophyll tissue (P: palisade, US: upper spongy mesophyll and UL: lower spongy mesophyll, all in mm2 mm-2).    37 3.3.5. Chloroplasts in mesophyll tissues (Chlo_P, Chlo_US) Regarding chloroplasts in the mesophyll, there were clearly many more in number and packed more closely together within the palisade tissue than within the spongy mesophyll (Figure 8). Several significant differences were found among genotypes, particularly for the palisade tissue (Table 11). This variation could reflect differences in either the size of cells or the density of chloroplasts within cells. Table 11. Analysis of variance for the count of chloroplasts, visible in cross-section, in palisade and upper spongy mesophyll (n=2) C Chlo_P ANOVA SD C Chlo_US ANOVA SD K4C8 10.767 a 0.841 K4C8 4.600 a 1.124 K4C1 11.833 ab 0.799 K4C18 5.100 ab 0.657 K4C6 12.000 abc 2.286 K4C7 5.183 abc 0.662 K4C2 12.350 abcd 1.305 K4C11 5.500 abcd 0.620 K4C7 12.367 abcde 0.831 K4C9 5.617 abcde 0.504 K4C15 12.517 abcdef 1.617 K4C17 5.650 abcdef 0.414 K4C9 12.567 abcdefg 1.164 K4C19 5.650 abcdefg 0.625 K4C3 12.633 bcdefgh 0.674 K4C20 5.667 abcdefgh 0.807 K4C18 12.700 bcdefghi 1.587 K4C14 5.850 abcdefghi 1.069 K4C19 12.767 bcdefghij 1.340 K4C13 6.117 abcdefghij 0.960 K4C11 13.200 bcdefghijk 1.124 K4C1 6.450 bcdefghijk 0.582 K4C14 13.283 bcdefghijkl 1.778 K4C12 6.467 bcdefghijkl 1.035 K4C13 13.550 bcdefghijklm 1.069 K4C4 6.800 cdefghijklm 0.955 K4C12 13.850 defghijklmn 1.211 K4C15 6.833 cdefghijklmn 0.809 K4C17 13.950 defghijklmno 1.492 K4C3 6.900 defghijklmno 0.782 K4C20 14.067 defghijklmnop 1.998 K4C6 6.933 defghijklmnop 0.463 K4C4 14.917 klmnop 1.254 K4C2 7.283 efghijklmnop 0.884 C: clones, SD: standard deviation, Chlo_P: genotypic means of counts of chloroplasts/cells in palisade, Chlo_US: genotypic means of counts of chloroplasts/cells in upper spongy mesophyll.  3.3.6 Nitrogen density per unit mesophyll wall area Na/Smes/S Nitrogen density divided by Smes/S_T is of interest because mesophyll conductance might be expected to vary inversely with the nitrogen content per unit area of cell wall space  38 available for CO2 diffusion. Significant differences, however, were not found for Na/Smes/S (Figure 9).  Figure 9. Nitrogen content per unit Smes/S_T, by clone. Error bars are ±SD. n = 2 for K4C-1, 6 and 13 (no error bars); n = 3 for K4C12; n = 4 for K4C-2, 8, 9, 11, 17, 20 and n= 5 for the rest.   00.010.020.030.040.050.060.070.080.09K4C1 K4C2 K4C3 K4C4 K4C6 K4C7 K4C8 K4C9 K4C11 K4C12 K4C13 K4C14 K4C15 K4C17 K4C18 K4C19 K4C20Na/Smes/SClones 39 3.4. Ultrastructure Despite high costs associated with its use, especially if done with a diamond knife, TEM is a very precise method for anatomical studies. The quality of the images obtained permitted easy identification of many cell components (Figure 10).  Figure 10. Transmission electron micrograph of K4C1 with all the specifications.   40 3.4.1. Cell wall thickness (tcw) The observed cell wall had a very homogenous thickness, but did vary among biological replicates, which explains, in part, the high variability evident in Figure 11. In combination, high standard deviations and the low number of biological replicates (just two per genotype) precluded finding any significant differences.  Figure 11.  Cell wall thickness averages, by clone. n = 2.  3.4.2. Thylakoid grana thickness (tgra) The thickness of the thylakoid grana was highly variable (Figure 12), and there were large differences between biological replicates. Similar to cell wall thickness, there were no significant differences found between genotypes in this trait.  00.050.10.150.20.250.30.350.4K4C1 K4C2 K4C3 K4C4 K4C6 K4C7 K4C8 K4C9 K4C11K4C12K4C13K4C14K4C15K4C17K4C18K4C19K4C20t cw(um)Clones 41  Figure 12.  Thylakoid grana thickness averages, by clone. n = 2.  3.4.3. Chloroplast surface area exposed to intercellular air space (Sc/S) This is the cell wall area adjacent to chloroplasts, expressed per unit leaf surface area. Despite the variation in this trait (Figure 13), there were no significant differences between genotypes.  Figure 13.  Chloroplast surface area exposed to intercellular air space per unit of leaf area, by clone. n = 2. 00.050.10.150.20.25K4C1 K4C2 K4C3 K4C4 K4C6 K4C7 K4C8 K4C9 K4C11 K4C12 K4C13 K4C14 K4C15 K4C17 K4C18 K4C19 K4C20t gra (um)Clones0.005.0010.0015.0020.0025.0030.00K4C1 K4C2 K4C3 K4C4 K4C6 K4C7 K4C8 K4C9 K4C11K4C12K4C13K4C14K4C15K4C17K4C18K4C19K4C20S c/S (mm2m-2)Clones 42 3.5. Stomata Stomata (Figure 14) are the gate between leaves and atmosphere for most CO2 and water vapor diffusion. This diffusion is facilitated by the stomatal conductance, gs, which may depend on stomatal density and the pore dimensions (length and width, the latter being controlled by stomatal behavior).  Figure 14. Stomatal impressions for genotypes K4C14, K4C18 and K4C20  43 The genotypes with the higher densities of stomata appeared to be K4C7, K4C13, K4C14 and K4C17. Statistical differences between genotypes were found in both, stomatal density and stomatal aperture (Table 12). The length of the stomatal aperture was very uniform among all genotypes. Table 12. Analysis of variance for the stomatal count per mm2 (n=2-5) C stoD ANOVA SD C stoLA ANOVA SD K4C8 193.6580 a 31.512 K4C19 18.409   3.794 K4C18 226.6141 ab 38.582 K4C17 19.310 a 3.163 K4C9 228.1283 abc 43.768 K4C14 19.584 b 4.332 K4C12 231.8617 abcd 49.981 K4C20 20.187 abc 2.253 K4C19 233.0610 abcde 156.497 K4C7 20.367 abcd 3.697 K4C20 233.0610 abcdef 75.794 K4C8 20.542 abcde 3.048 K4C15 235.9262 abcdefg 30.293 K4C18 20.699 cdef 2.539 K4C11 254.4367 abcdefgh 58.674 K4C11 20.991 cdefg 3.563 K4C6 268.9103 abcdefghi 20.540 K4C6 21.139 cdefgh 4.162 K4C7 277.3589 abcdefghij 54.691 K4C15 21.515 efghi 2.883 K4C13 334.1436 bcdefghijk 70.574 K4C13 21.677 defghij 3.060 K4C14 378.7898 ikl 130.656 K4C12 21.913 fghijk 3.401 K4C17 385.4874 ikl 54.919 K4C9 21.940 efghijk 2.832 C: clones, SD: standard deviation, stoD: genotypic means of the Stomatal density; stoLA: genotypic means of the length of the stomatal aperture.  3.6. Trait correlations For purposes of detecting genotypic differences, there was a reasonable sample size for most of the tissue measurements but often not for the microscopic characters, particularly those that depended on electron microscopy. Several traits, such as C:N, LMA, Na/Smes/S and Na, did not show any significant difference through ANOVA. Correlation analysis across the whole dataset provided an additional approach to explore the measured properties and their relationships to each other. Table 13 presents a correlation matrix for key traits that may impact gm.   44 Table 13. Pearson’s product-moment correlation coefficients (r). Colors from light to dark indicate significance levels of 0.05, 0.01 and 0.001.  LPI CCI LMA tleaf tpal tspo fias Pmes/S_P Pmes/S_US Pmes/S_LS Pmes/S_T Smes/S_P Smes/S_US Smes/S_LS Smes/S_T Na Na/Smes/S Sc/S Chlo_P Chlo_US tcw CCI 0.715                     LMA 0.365 0.718                    tleaf 0.615 0.596 0.515                   tpal 0.540 0.692 0.700 0.602                  tspo 0.471 0.497 0.498 0.871 0.328                 fias -0.105 -0.129 -0.510 -0.019 -0.152 -0.229                Pmes/S_P 0.576 0.707 0.813 0.546 0.563 0.514 -0.497               Pmes/S_US 0.447 0.473 0.409 0.427 0.150 0.552 -0.445 0.753              Pmes/S_LS 0.287 0.239 0.307 0.613 0.081 0.791 -0.313 0.515 0.701             Pmes/S_T 0.524 0.584 0.624 0.583 0.345 0.663 -0.493 0.904 0.936 0.776            Smes/S_P 0.393 0.356 0.581 0.329 0.566 0.208 -0.189 0.422 -0.035 -0.117 0.154           Smes/S_US 0.627 0.567 0.395 0.491 0.288 0.496 -0.308 0.760 0.916 0.541 0.866 0.099          Smes/S_LS 0.309 0.226 0.292 0.630 0.081 0.815 -0.331 0.426 0.639 0.937 0.694 -0.092 0.542         Smes/S_T 0.677 0.592 0.657 0.691 0.511 0.694 -0.399 0.809 0.708 0.577 0.812 0.597 0.792 0.616        Na 0.236 0.742 0.841 0.288 0.574 0.245 -0.296 0.631 0.242 0.015 0.398 0.375 0.299 0.039 0.390       Na/Smes/S -0.555 -0.225 -0.188 -0.481 -0.288 -0.520 0.389 -0.524 -0.618 -0.528 -0.629 -0.302 -0.685 -0.545 -0.746 0.109      Sc/S 0.490 0.473 0.469 0.156 0.313 0.198 -0.280 0.606 0.571 0.333 0.600 0.443 0.554 0.286 0.658 0.330 -0.394     Chlo_P 0.281 0.502 0.549 0.083 0.488 0.050 -0.278 0.396 0.125 -0.165 0.196 0.610 0.110 -0.155 0.352 0.478 -0.113 0.462    Chlo_US 0.316 0.334 0.225 0.066 -0.106 0.169 -0.083 0.475 0.603 0.418 0.573 -0.032 0.589 0.437 0.459 0.239 -0.146 0.634 0.196   tcw 0.156 0.401 0.264 0.279 0.410 0.073 0.101 0.241 0.011 -0.097 0.093 0.121 0.233 -0.039 0.178 0.327 -0.121 -0.252 0.119 -0.047  tgra 0.195 0.261 0.133 0.207 0.200 0.081 -0.142 0.333 0.427 0.188 0.378 -0.082 0.559 0.263 0.348 0.092 -0.447 0.090 0.014 0.225 0.714  LPI: leaf plastochron index; CCI: chlorophyll content index; LMA: leaf mass per unit area; tleaf: leaf thickness; tpal: palisade thickness; tspo: spongy mesophyll thickness; fias: fraction of mesophyll occupied by intercellular airspace; Pmes/S_T: total surface area of mesophyll cells, per unit area of leaf surface, calculated from total perimeter of cell walls; Pmes/S_P: surface area of mesophyll cells calculated from total perimeter of cell walls in palisade tissue; Pmes /S_US: surface area of mesophyll cells calculated from total perimeter of cell walls in upper spongy mesophyll; Pmes/S_LS: surface area of mesophyll cells calculated from total perimeter of cell walls in lower spongy mesophyll; Smes/S_P: surface area of mesophyll cells exposed to airspace in palisade tissue; Smes/S_US: surface area of mesophyll cells exposed to airspace in upper spongy mesophyll; Smes/S_LS : surface area of mesophyll cells exposed to airspace in lower spongy mesophyll; Smes/S_T: total surface area of mesophyll cells exposed to airspace, per unit of leaf area Na: nitrogen per unit leaf area; Na/Smes/S: nitrogen per unit mesophyll area; Sc/S: chloroplast surface area exposed to intercellular airspace per unit of leaf area; Chlo_P: average count of visible chloroplasts per  palisade cells; Chlo_US: average count of visible chloroplasts per  upper spongy mesophyll cells; tcw: cell wall thickness; tgra: thylakoid grana thickness.  45 There were significant correlations between LPI, CCI and measures of leaf thickness and, often, cell wall surface areas. Nitrogen per unit exposed mesophyll cell wall area was negatively correlated with LPI. The LPI did not correlate with nitrogen density but, as would be expected, CCI did. Likewise, CCI and Na were highly correlated with LMA, but LMA and LPI were not. Several correlations were expected to be high where the traits depended, in part, on the same or related measurements. For example, thickness of the leaf was correlated with thickness of the component layers of the mesophyll, which are correlated with each other. Similarly, Pmes/S_T and Smes/S_T, and most of their component areas, co-varied extensively, amongst themselves and with tleaf and its components. Hence, thicker leaves present a greater mesophyll cell wall surface area. In addition, this also translates into a strong positive correlation between LMA and both Pmes/S_T and Smes/S_T, in addition tleaf (Table 13). These correlations with LMA are of particular importance to the present work in order to understand the basis of variation in this trait. The single best anatomical predictor of LMA was surface area of mesophyll cells in palisade tissue (Pmes/S_P), which accounts for well over half of the total cell wall area of the mesophyll (Table 8). The relationship between Pmes/S_P and LMA is described by the following regression equation (r=0.813, p<0.001):  LMA = 0.0128 + (0.00112 × Pmes/S_P) [5]  46 Besides tissue thickness, Pmes/S_P (and, more generally, Pmes/S_T) is expected to vary with cell packing, the reverse of fias. Consistent with this, fias was negatively correlated with Pmes/S_P and Pmes/S_T. Likewise, fias was negatively correlated with LMA. Genotypic variation in LMA as a function of both tleaf and fias is shown in Figure 15, and the following multiple regression equation describes their independent, but equally significant effects (r=0.718, adjusted r2=0.466; p=0.018 for fias, p=0.017 for tleaf):  LMA = 0.0512 – (0.0541 × fias) + (0.0953 × tleaf) [6]   Figure 15. Genotypic variation in LMA as a function of both tleaf and fias. Looking at these two components of the LMA, is possible to see that in thicker leaves, LMA goes up. At the same time, high cell packing (i.e., low fias) is also related to high LMA.  47 Nitrogen content expressed per unit exposed mesophyll surface area (Na/Smes/S) was not related to LMA, but decreased significantly with leaf thickness, and particularly the thickness of the spongy mesophyll. Note that negative correlations between Na/Smes/S and measures of cell wall surface area may be spurious because of autocorrelation, and need to be interpreted cautiously. Regarding chloroplast counts, there were some positive correlations between area measurements and either Chlo_P and Chlo_US, in general accordance with expectations (e.g., a higher count of chloroplasts in palisade tissue was associated with higher Smes/S_P, and similarly for chloroplasts in upper spongy mesophyll and Smes/S_US). Interestingly, cell wall thickness and granum thickness were well correlated (Table 13).  Granum thickness also showed some correlation with Smes/S_US, but cell wall thickness did not correlate with any other measure. Stomatal density (stoD) and aperture length (stoLA) (not shown) did not correlate with any other traits, but were negatively associated with each other.          48 CHAPTER 4. Discussion The aim of the present study was to conduct an extensive descriptive analysis of anatomical variables that may limit or influence gm in leaves of balsam poplar. To the best of my knowledge, this is the first and only study examining leaf ultrastructure in this species and one of just few studies looking at leaf structure in balsam poplar (Al Afas et al. 2007; Castro and Sanchez-Azofeifa 2008). Though restricted to a single, but highly variable family, this extensive work will assist future researchers in understanding how structure can constrain function. It is possible to appreciate from the data coming from the light micrographs that variation within this family was not solely in terms of leaf thickness; rather, differences reflected complex interactions of cell size, cell orientation, cell association within tissue, leaf mass, leaf area and so forth.  4.1. Study limitations In the case of higher plants, the leaves emerge periodically from the shoots. Consequently, the period between consecutive leaves is described as the plastochron (Erickson and Michelini 1957). Although an effort was made to sample leaves of uniform developmental stage (i.e., visibly, the youngest fully mature leaf), I calculated the Leaf Plastochron Index (LPI) of said leaves and found that it varied between genotypes. It was anticipated that leaves with LPI of six or seven would have a structure that should not change, because at this point everything is already built. Indeed, LPI values found in the literature support this statement. For instance, Populus tremula leaves are considered mature when LPI ≥ 5, while values of LPI < 5 are associated with the continuous changes in leaf anatomy expected of immature leaves (Tosens et al. 2012). In the case of hybrid  49 poplar clones, an LPI of 12 or more defines mature leaves (Théroux-Rancourt et al. 2014). Leaves sampled in the present study mostly fell between these values and it is clear from the correlation matrix (Table 13) that leaf properties may have changed over this range. Several studies have mentioned the importance of working with the appropriate developmental stage of the plant material for a given study (Eichelmann et al. 2004; Tosens et al. 2012; Slaton and Smith 2002). The reason is to avoid complications caused by the increasing leaf photosynthetic capacity of developing leaves, expressed on both a per unit area and dry mass basis, as well as the accumulation of Rubisco (Eichelmann et al. 2004).  Leaf development has been highly related to increases in “dry mass per unit area, thickness, density, exposed mesophyll (Smes/S), and chloroplast (Sc/S) to leaf area ratio, internal air space (fias), cell wall thickness and chloroplast dimensions” (Tosens et al. 2012). In this study, LPI had high correlations (Table 13) with some of the above-described traits, such as tleaf (r=0.615, p<0.01) and Smes/S_T (r=0.677, p<0.01). Foremost, the highest correlation was seen between LPI and CCI (r=0.715, p<0.001). Consequently, leaf developmental stage may have influenced genotypic comparisons among the other traits. Another limitation of this study was the low sample size, particularly for traits measured microscopically. Efforts were taken to ensure sufficient biological replicates for each clone, but there were some losses in the greenhouse and, subsequently, losses during sample preparation for microscopy (resin infiltration issues). Furthermore, it would have  50 been difficult to accommodate more samples, given the labor required. In retrospect, it might have been better to sample fewer genotypes more intensively, with more biological replicates but fewer technical replicates. Despite limitations in sample size, I was able to detect genotypic variation in several traits, and correlations between traits across genotypes. 4.2. Whole tissue properties The C:N ratio is reflective of the nitrogen concentration of plant tissue and is often taken to be an indicator of nutritional status (Kruse et al. 2010). The average C:N ratio for data presented in Figure 6 was 9.5, which is similar to values reported for N-sufficient hybrid poplar (Kruse et al. 2003), and well below (i.e., better than) a sufficiency range (16-20) reported for Populus species in general (Bergmann 1992). The foliar C:N ratios of field P. trichocarpa (grown for 3 years) were close to 21 (McKown et al. 2014a).  Nitrogen is essential for the production of thylakoid proteins, chlorophylls, and enzymes (Wright et al. 2004a). In fact, more than 75% of leaf organic nitrogen is used for this purpose, Rubisco accounting for 20-40% (Evans 2014). As previously mentioned, northern populations within the species increase their photosynthetic capacity, due to higher Na (Soolanayakanahally 2010), which can be accomplished through either a higher N concentration (mg mg-1 DW) or higher LMA. The genotypes in the present study did not differ in C:N ratio, N concentration (not presented), Na or LMA, but LMA and Na were very highly correlated (r=0.841, p<0.001) across genotypes (Table 13). Leaf greenness (i.e., CCI) is often used as a proxy measure of foliar N content. Chlorophyll content index showed substantial variation among genotypes (Table 3) and,  51 as expected, was highly correlated with Na (r=0.742, p<0.001) and LMA (r=0.718, p<0.01).  4.3. Microanatomy Leaf thickness is central to the three hypotheses of this study, where thickness is thought to influence LMA, gm and N concentration. Leaf thickness will depend on the thickness of the ground tissue (i.e., the mesophyll) plus the epidermal layers. In the balsam poplar leaves used in this study, the mesophyll consistently had two well-packed layers of elongated palisade cells with two clear zones of spongy mesophyll below. There were significant genotypic differences in tleaf, tpal and tspo (Table 4). The thicknesses of the palisade and the spongy mesophyll were roughly similar but there was greater variation in the latter. Consequently, tleaf was better correlated with the thickness of the spongy mesophyll than with the thickness of the palisade (r=0.871, p<0.001 and r=0.602, p<0.05, respectively). In corollary, LMA was better correlated with tpal (r=0.700, p<0.01) than with tleaf (r=0.515, p<0.05). Visually, the lower spongy mesophyll seemed highly variable; most of the cells had very dissimilar shapes, were positioned individually, and were not clustered as in the upper spongy mesophyll and palisade tissues. Likewise, the cell wall area (Pmes/S) of the spongy mesophyll, and particularly the lower spongy mesophyll, was more variable than the palisade tissue (Table 9). However, because of the high cross-sectional area of palisade tissue as well as the high surface area to volume ratio of palisade cells, more than half of the cell wall area (and usually more than half of the exposed wall area, Smes/S) was in the  52 palisade. The ratio of cell wall area in the palisade relative to the upper and lower spongy mesophyll was approximately 4:2:1 when averaged across genotypes. Given that most of the cell wall area is in the palisade, and considering that this area also closely represents the surface area and likely the volume of the cytosol, it is not surprising that LMA was well correlated with Pmes/S_T (r=0.624, p<0.01) and especially with Pmes/S_P (r=0.813, p<0.001). LMA was also correlated with Smes/S_T (r=0.657, p<0.01) and Smes/S_P (r=0.581, p<0.05), but less so because of the removal of a varying amount of adjoining cell wall area from these measures.   Smes/S_T describes the actual cell wall area available for CO2 diffusion in photosynthesis. The mean value across all balsam poplar genotypes was 32 mm mm-2. This is approximately half the value of 78 mm mm-2 that was reported by Terashima et al. (2006) for Japanese poplar (Populus maximowiczii A. Henry). There was a strong correlation between Smes/S_T and tleaf in balsam poplar (r=0.691, p<0.01), similar to what has been reported by other workers for a variety of species (Slaton and Smith 2002; Giuliani et al. 2013). Many studies that have focused on leaf thickness are related to comparisons between sun and shade leaves, and how these differences relate to the photosynthetic performance of species (Bergen 1904; Yano and Terashima 2001; Miyazawa and Hanba 2001; Evans and Poorter 2001; Buckley and Warren 2014; Garnier et al. 1999; Wright et al. 2004b). Sun leaves are much thicker than shade leaves. Sun leaves also have much greater amounts of Rubisco per unit leaf area, and greater Smes/S (and larger Sc/S) than shade leaves. Moreover, this is accompanied by a higher gm in sun leaves compared to shade leaves (Hanba et al. 2002).   53 In addition to leaf thickness, LMA, gm and N concentration will depend on tissue density; that is, the packing of cells and the fraction of the mesophyll volume that is air space (fias). Given a constant thickness, any reduction in fias would be expected to increase LMA, as well as the tortuosity of the diffusion pathway from the substomatal cavities to the chloroplasts. Indeed, there was a negative correlation between fias and LMA (r=-0.510, p<0.05). A reduction in fias must, by definition, be accompanied by an increase in the cell fraction (i.e., 1-fias) which should yield an increase in cell wall surface area, as reflected by negative correlations between fias and Pmes/S_P (r=-0.497, p<0.05) and Pmes/S_T (r=-0.493, p<0.05). An increase in cell wall area should positively affect gm if the additional surface is exposed to intercellular airspace, but fias and Smes/S_T were not correlated. At some point, an increased cell fraction must also lead to increased cell contact, which would negatively impact gm. 4.4. Ultrastructure In terms of ultrastructure, cell wall thickness and chloroplast distribution are known to be the most important factors affecting gm (Buckley and Warren 2014; Scafaro et al. 2011; Miyazawa and Terashima 2001; Terashima et al. 2011). In fact, Terashima et al. (2011) demonstrated that cell wall resistance is responsible for half of the total mesophyll resistance to CO2 diffusion. Most of the resistance to gm in the liquid phase is thought to be a function of cell wall thickness (tcw) and cell wall surface area exposed to intercellular airspace (Smes/S) or, more specifically, chloroplast surface area exposed to airspace (Sc/S). In some species, and particularly those with sclerophytic leaves (Tomás et al. 2013), tcw is more important than Sc/S. Along the same lines, Scafaro et al. (2011) found that thicker mesophyll cell walls in wild rice species are likely responsible for a reduction in gm,  54 which was associated with a greater drawdown of CO2 into chloroplasts (Ci–Cc) compared to domesticated Oryza sativa. I did not detect any variation in tcw between the balsam poplar genotypes used in this study. The cell wall is responsible for support and protection of all cells; hence, it must have a rigid and firm structure. As noted earlier, there is a strong relationship between cell wall area and LMA. If cell wall mass contributes significantly to LMA, then one might also expect tcw to correlate with LMA, but no such relationship was found. The cell wall thickness was, however, well correlated with the thylakoid grana thickness (r=0.714, p<0.01). The thylakoids are the site for the light-dependent reactions of photosynthesis, and there is an inverse relationship between photosynthetic capacity, light availability and the thickness of the grana inasmuch as shade leaves/chloroplasts typically have thicker grana than sun leaves/chloroplasts (Lichtenthaler et al. 1981, Yano and Terashima 2001). Shade leaves also have a lower photosynthetic capacity and thus may not require high gm. It’s possible that a similar relationship drives the genotypic correlation between tcw and tgra in balsam poplar. Regarding Sc/S and gm, it has been established that there is a strong positive relationship across species belonging to the same functional groups (e.g., annuals, broad leaved trees, etc.) (Terashima et al. 2006; Tomás et al. 2013). Although I found significant genotypic variation in all of Pmes/S_P, Pmes/S_US, Pmes/S_LS, Pmes/S_T, Smes/S_P, Smes/S_US, Smes/S_LS, and Smes/S_T, there were no apparent differences in Sc/S (Figure 12). This may be because I was working with just one species and/or because Sc/S is a very sensitive trait that is directly influenced by local light conditions (Terashima et al. 2006). Nonetheless, Sc/S was positively correlated with LMA (r=0.469, p<0.05) and also with  55 measures of palisade, upper spongy mesophyll and total cell wall area, whether exposed to intercellular air space or not. Assuming that most foliar nitrogen is involved in photosynthesis, the leaf nitrogen per unit exposed mesophyll surface area (Na/Smes/S) might be taken as a measure of the relative wall resistance (i.e., the inverse of the amount of wall area available for CO2 diffusion per unit N). This did seem to go down with tleaf and tspo (Table 13), but was unrelated to LMA. 4.5. Stomata Prior to reaching the mesophyll, CO2 must first diffuse into the leaf through the stomata.  Although the stomatal conductance (gs) is distinct from gm, certain aspects of stomatal anatomy can influence gm. For example, the placement of stomata on both leaf surfaces (amphistomy) can mitigate the effect of leaf thickness on gm by decreasing the mean diffusion path-length through the mesophyll (Muir et al. 2014). Adaxial stomata are rare in balsam poplar (Soolanayakanahally et al. 2009), but stomatal density (stoD) on the abaxial surface might also influence path-length. It remains to be seen whether stoD has any relationship to gm, but stoLA was negatively correlated (r=-0.646, p<0.01) to stoD in the present study. Ignoring stomatal behavior, if stoD were to increase without an accompanying increase in gs, then length of the stomatal aperture (stoLA) would have to decrease.    56 4.6. Structural basis of variation in LMA One of the key variables in the “leaf economic spectrum” is LMA (Wright et al. 2004b). Species with high LMA may have thicker leaves or denser tissues or both. As noted in the introduction, LMA has been associated both positively and negatively with gm in a variety of studies. For example, northern balsam poplar genotypes seem to have both high gm and LMA (Soolanayakanahally et al. 2009); whereas in wild tomatoes (Lycopersicon sp.), Muir et al. (2014) reported that greater LMA led to reduced gm because of increased cell packing and leaf thickness, together. Increased cell packing is reflected by a decrease in fias. In the present study, both fias and leaf thickness (tleaf) were equally correlated with LMA. Of the structural variables investigated, the palisade cell wall surface area (Pmes/S_P) was best correlated with LMA, as described by Equation 5. However, this variable itself reflects the combined effects of thickness (increasing the length of the palisade cells) and cell packing (increasing the numbers of palisade cells). The combination of these more fundamental variables, tleaf and fias, can account for considerable variation in LMA within the K4×C balsam poplar family (Equation 6, Figure 15). 4.7. Correlation with gm  Given that LMA is at least partly related to leaf thickness, it is counterintuitive that gm would increase with LMA because of the negative effect on the diffusion path-length. Alternatively, the observed increase in mesophyll cell wall surface area servicing photosynthesis (Sc/S) should enhance gm. Hence, a straightforward relationship between LMA and gm is not expected.  57 It was not possible within the scope of the present work to measure gm on the specific leaves I sampled. However, two co-workers have independently explored genotypic variation in gm in the K4×C family using two different methods. Based on an analysis of CO2 response curves, Natalie Ryan (MSc thesis, in preparation) found a curvilinear relationship between gm and LMA, the former increasing with LMA to a plateau. However, using chlorophyll fluorescence techniques, Mina Momayyezi (in progress) detected no relationship between gm and LMA. The first dataset was unavailable at the time of this writing. Correlation analysis between the second dataset and anatomical traits measured in this thesis (Appendix) revealed little. It’s important to emphasize that measurement of gm is difficult and very prone to error. This, coupled with the fact that gm has both fixed and variable components, and only the former are structural, means that a proper assessment of the relationship between gm and the leaf traits measured here awaits a more thorough physiological assessment of the K4×C family. A focus on the extreme genotypes for particular traits identified in this thesis may facilitate this comparison.   4.8. Conclusions As hoped, this study revealed large differences in leaf anatomy within the K4×C family. The different siblings present different combinations of traits that may be useful in approaching hypotheses related to form and function. Considering the significant differences and genotypic rank orders found for several key traits (e.g., tleaf, Pmes/S, Smes/S, fias), the most extreme candidates for high vs low gm would appear to be K4C13 and  58 K4C8, respectively. These genotypes also had or were near the extreme values for LMA and Sc/S, and would therefore be the first choices for comparative physiological work. The first hypothesis of this study, h1: “LMA varies directly with thickness of the leaf, and is a suitable proxy measure of such”, is accepted, but only in part because cell packing also contributes significantly to LMA in balsam poplar. For any given genotype, a high LMA may reflect high tleaf, low fias, or both (Figure 15). The second hypothesis, h2: “Leaves with higher LMA will have more cell wall area for CO2 diffusion to chloroplasts, or lower nitrogen per unit exposed cell wall area”, is also accepted in part. As indicated by the trait correlation analysis (Table 13), cell wall surface area (Pmes/S) varied directly with LMA, and had the strongest relationship with LMA. Smes/S and Sc/S were also positively correlated with LMA. Nitrogen per unit exposed cell wall area (Na/Smes/S) was not correlated with LMA, but did go down as tleaf and tspo increased. Essentially, when leaves get thicker and the cell wall area increases, N content is spread out inside the leaf. Unfortunately and given the limitations of this study, the third hypothesis, “h3: gm varies as a function of leaf microanatomy and/or ultrastructure” could not be properly tested. It is for the mean time rejected until additional functional data become available.  Future studies should link these findings, and both gm and gs, in research performed simultaneously on the same tissue, to gain insights in the photosynthetic capacity of this species. The present study provides guidance as to which traits are most worth pursuing, and identifies particular genotypes that may be targeted in those investigations.   59 Bibliography Al Afas N, Marron N, Ceulemans R. 2007. Variability in Populus leaf anatomy and morphology in relation to canopy position, biomass production, and varietal taxon. Annals of Forest Science, 64, 521–532. Bergen JY. 1904. Transpiration of sun leaves and shade leaves of Olea europaea and other broad-leaved evergreens. Botanical Gazette, 38, 285–296. Bergmann W. 1992. Nutritional Disorders of Plants: Visual and Analytical Diagnosis. Gustav Fischer, Jena, Germany. 386 pp.  Buckley TN, Warren CR. 2014. The role of mesophyll conductance in the economics of nitrogen and water use in photosynthesis. Photosynthesis Research, 119, 77–88. Castro KL, Sanchez-Azofeifa GA. 2008. Changes in spectral properties, chlorophyll content and internal mesophyll structure of senescing Populus balsamifera and Populus tremuloides leaves. Sensors, 8, 51–69. Chabot BF, Chabot JF. 1977. Effects of light and temperature on leaf anatomy and photosynthesis in Fragaria vesca. Oecologia, 26, 363–377. Cormack RGH. 1950. A study of leaf thickness in wheat. Agronomy Journal, 42, 361–362. Eichelmann H, Oja V, Rasulov B, Padu E, Bichele I, Pettai H, Niinemets Ü, Laisk A. 2004. Development of leaf photosynthetic parameters in Betula pendula Roth leaves: correlations with photosystem I density. Plant Biology, 6, 307–318. Erickson RO, Michelini FJ. 1957. The plastochron index. American Journal of Botany, 44, 297–305. Evans JR. 1983. Nitrogen and photosynthesis in the flag leaf of wheat Triticum aestivum L.. Plant Physiology, 72, 297–302. Evans JR. 1989. Photosynthesis and nitrogen relationships in leaves of C3 plants. Oecologia, 78, 9–19. Evans JR, von Caemmerer S, Setchell BA, Hudson GS. 1994. The relationship between CO2 transfer conductance and leaf anatomy in transgenic tobacco with a reduced content of Rubisco. Australian Journal of Plant Physiology, 21, 475–495. Evans JR, Kaldenhoff R, Genty B, Terashima I. 2009. Resistances along the CO2 diffusion pathway inside leaves. Journal of Experimental Botany, 60, 2235–2248.   60 Evans JR, von Caemmerer S. 1996. Carbon dioxide diffusion inside leaves. Plant Physiology, 339–346. Evans JR, Poorter H. 2001. Photosynthetic acclimation of plants to growth irradiance: the relative importance of specific leaf area and nitrogen partitioning in maximizing carbon gain. Plant, Cell and Environment, 24, 755–767. Flexas J, Ribas-Carbó M, Diaz-Espejo A, Galmés J, Medrano H. 2008. Mesophyll conductance to CO2: current knowledge and future prospects. Plant, Cell and Environment, 31, 602–621. Flexas J, Barbour M, Brendel O, Cabrera H, Carriquí M, Díaz-Espejo A, Douthe C, Dreyer E, Ferriog J, Gago J, Gallé A, Galmés J, Kodama N, Medrano H, Niinemets Ü, Peguero-Pina JJ, Pou A, Ribas-Carbó M, Tomás M, Tosens T, Warren C. 2012. Mesophyll diffusion conductance to CO2: an unappreciated central player in photosynthesis. Plant Science, 193-194, 70–84. Friend DJC. 1961. The control of chlorophyll accumulation in leaves of marquis wheat by temperature and light intensity. II. Chlorophyll contents relative to leaf area and thickness. Physiologia Plantarum, 14, 28–39. Garnier E, Salager JL, Laurent G, Sonié L. 1999. Relationships between photosynthesis, nitrogen and leaf structure in 14 grass species and their dependence on the basis of expression. New Phytologist, 143, 119–129. Giuliani R, Koteyeva N, Voznesenskaya E, Evans MA, Cousins AB, Edwards GE. 2013. Coordination of leaf photosynthesis, transpiration, and structural traits in rice and wild relatives (Genus Oryza). Plant Physiology, 162, 1632–1651. Hanba YT, Miyazawa SI, Kogami H, Terashima I. 2001. Effects of leaf age on internal CO2 transfer conductance and photosynthesis in tree species having different types of shoot phenology. Australian Journal of Plant Physiology, 28, 1075–1084. Hanba YT, Kogami H, Terashima I. 2002. The effect of growth irradiance on leaf anatomy and photosynthesis in Acer species differing in light demand. Plant, Cell and Environment, 25, 1021–1030. Hanba YT, Miyazawa SI, Terashima I. 2014. The influence of leaf thickness on the CO2 transfer conductance and leaf stable carbon isotope ratio for some evergreen tree species in Japanese warm-temperate forests. Functional Ecology, 13, 632–639. Kaldenhoff R. 2012. Mechanisms underlying CO2 diffusion in leaves. Current Opinion in Plant Biology, 15, 276–281.  61 Kozak A, Kozak RA, Staudhammer CL, Watts SB. 2008. Introductory Probability and Statistics: Applications for Forestry and the Natural Sciences. CAB International, Wallingford, Oxfordshire, UK. 448 pp.  Kruse J, Hetzger I, Mai C, Polle A, Rennenberg H. 2003. Elevated pCO2 affects N-metabolism of young poplar plants (Populus tremula × Populus alba) differently at deficient and sufficient N-supply. New Phytologist, 1, 65–81.  Kruse J, Hänsch R, Mendel RR, Rennenberg H. 2010. The role of root nitrate reduction in the systemic control of biomass partitioning between leaves and roots in accordance to the C/N-status of tobacco plants. Plant Soil, 332, 387–403. Lichtenthaler HK, Buschmann C, Doll M, Fietz HJ, Bach T, Kozel U, Meier D, Rahmsdorf U. 1981. Photosynthetic activity, chloroplast ultrastructure and leaf characteristics of high-light and low-light plants and of sun and shade leaves, Photosynthesis Research, 2, 115–141. Loreto F, Di Marco G, Tricoli D, Sharkey TD. 1994. Measurements of mesophyll conductance, photosynthetic electron transport and alternative electron sinks of field grown wheat leaves. Photosynthesis Research, 41, 397–403. McKown AD, Guy RD, Klápště J, Geraldes A, Friedmann M, Cronk QCB, El-Kassaby YA, Mansfield SD, Douglas CJ. 2014a. Geographical and environmental gradients shape phenotypic trait variation and genetic structure in Populus trichocarpa. New Phytologist, 201, 1263–1276. McKown AD, Guy RD, Quamme L, Klápště J, La Mantia J, Constabel, CP, El-Kassaby YA, Hamelin RD, Zifkin M, Azam MS. 2014b. Association genetics, geography and ecophysiology link stomatal patterning in Populus trichocarpa with carbon gain and disease resistance trade-offs. Molecular Ecology 23, 5771–5790.  Meidner H. 1953. An instrument for the continuous determination of leaf thickness changes in the field. Journal of Experimental Botany 3, 319–325. Miyazawa SI, Hanba YT. 2001. Why are sun leaves thicker than shade leaves? - Consideration based on analyses of CO2 diffusion in the leaf. Journal of Plant Research 114 93–105. Miyazawa SI, Terashima I. 2001. Slow development of leaf photosynthesis in an evergreen broad-leaved tree, Castanopsis sieboldii: Relationships between leaf anatomical characteristics and photosynthetic rate. Plant, Cell and Environment, 24, 279–291. Muir CD, Hangarter RP, Moyle LC, Davis PA. 2013. Morphological and anatomical determinants of mesophyll conductance in wild relatives of tomato (Solanum sect.  62 Lycopersicon, sect. Lycopersicoides; Solanaceae). Plant, Cell and Environment, 37, 1415-1426. Niinemets Ü. 1999. Research review: Components of leaf dry mass per area –- thickness and density – alter leaf photosynthetic capacity in reverse directions in woody plants. New Phytologist, 144,35–47. Niinemets Ü, Kull O, Tenhunen JD. 2004. Within-canopy variation in the rate of development of photosynthetic capacity is proportional to integrated quantum flux density in temperate deciduous trees. Plant, Cell and Environment, 27, 293–313. Parkhurst DF. 1994. Tansley Review of No 65. Diffusion of CO2 and other gases inside leaves. New Phytologist,126, 449–479. Peguero-Pina JJ, Flexas J, Galmés J, Niinemets Ü, Sancho-Knapik D, Barredo G, Villarroya D, Gil-Pelegrín E, 2012. Leaf anatomical properties in relation to differences in mesophyll conductance to CO2 and photosynthesis in two related Mediterranean Abies species. Plant, Cell and Environment, 35, 2121–2129. Poorter L, Bongers F. 2006. Leaf traits are good predictors of plant performance across 53 rain forest species. Ecology, 87, 1733–1743. Romero-Aranda R, Bondada BR, Syvertsen JP, Grosser JW. 1997. Leaf characteristics and net gas exchange of diploid and autotetraploid Citrus. Annals of Botany, 79, 153–160. Saarela JM, Gillespie LJ, Consaul LL, Bull RD. 2012. Balsam poplar (Populus balsamifera; Salicaceae) beyond the tree line in the western Canadian mainland arctic (Northwest Territories). Artic, 65, 1–12. Scafaro AP, von Caemmerer S, Evans JR, Atwell BJ. 2011. Temperature response of mesophyll conductance in cultivated and wild Oryza species with contrasting mesophyll cell wall thickness. Plant, Cell and Environment, 34, 1999–2008. Slaton MR, Smith WK. 2002. Mesophyll architecture and cell exposure to intercellular air space in alpine, desert, and forest species. International Journal of Plant Sciences, 163, 937–948. Soolanayakanahally RY, Guy RD, Silim SN, Drewes EC, Schroeder WR. 2009. Enhanced assimilation rate and water use efficiency with latitude through increased photosynthetic capacity and internal conductance in balsam poplar (Populus balsamifera L.). Plant, Cell and Environment, 32, 1821–1832. Soolanayakanahally RY. 2010. Latitudinal Gradients in Adaptive Traits of Populus, PhD thesis. The University of British Columbia, Vancouver, Canada. 131 pp.  63 Spurr AR. 1969. A low-viscosity epoxy resin embedding medium for electron microscopy. Journal of Ultrastructure Research, 26, 31–43. Terashima I, Hanba YT, Tazoe Y, Vyas P, Yano S. 2006. Irradiance and phenotype: comparative eco-development of sun and shade leaves in relation to photosynthetic CO2 diffusion. Journal of Experimental Botany, 57, 343–54. Terashima I, Hanba YT, Tholen D, Niinemets Ü. 2011. Leaf functional anatomy in relation to photosynthesis. Plant Physiology, 155, 108–16. Thain J.F. 1983. Curvature correction factors in the measurement of cell surface areas in plant tissues. Journal of Experimental Botany 34, 87–94. Théroux-Rancourt G, Ethier G, Pepin S. 2014. Threshold response of mesophyll CO2 conductance to leaf hydraulics in highly transpiring hybrid poplar clones exposed to soil drying. Journal of Experimental Botany, 65, 741–53. Tomás M, Flexas J, Copolovici L, Galmés J, Hallik L, Medrano H, Ribas-Carbó M, Tosens T, Vislap V, Niinemets Ü. 2013. Importance of leaf anatomy in determining mesophyll diffusion conductance to CO2 across species: quantitative limitations and scaling up by models. Journal of Experimental Botany, 64, 2269–2281. Tosens T, Niinemets Ü, Vislap V, Eichelmann H, Castro Díez P. 2012. Developmental changes in mesophyll diffusion conductance and photosynthetic capacity under different light and water availabilities in Populus tremula: How structure constrains function. Plant, Cell and Environment, 35, 839–856. Villar R, Ruiz-Robleto J, Ubera JL, Poorter H. 2013. Exploring variation in leaf mass per area (LMA) from leaf to cell: an anatomical analysis of 26 woody species. American Journal of Botany, 100, 1969–1980. Wilson PJ, Thompson K, Hodgson JG. 1999. Specific leaf area and leaf dry matter content as alternative predictors of plant strategies. New Phytologist, 143, 155–162. Wright, I., Groom, P., Lamont, B., Poot, P., Prior, D., Reich, P., Schulze, D., Veneklaas, E. and Westoby, M., 2004a. Short communication: Leaf trait relationships in Australian plant species. Functional Plant Biology, 31, 551-558. Wright IJ, Reich PB, Westoby M, Ackerly DD, Baruch Z, Bongers F, Cavender-Bares J, Chapin T, Cornelissen JHC, Diemer M, Flexas J, Garnier E, Groom PK, Gulias J, Hikosaka K, Lamont BB, Lee T, Lee W, Lusk C, Midgley JJ, Navas ML, Niinemets Ü, Oleksyn J, Osada N, Poorter H, Poot P, Prior L, Pyankov VI, Roumet C, Thomas SC, Tjoelker MG, Veneklaas EJ, Villar R. 2004b. The worldwide leaf economics spectrum. Nature, 428, 821–827.  64 Yano S, Terashima I. 2001. Separate localization of light signal perception for sun or shade type chloroplast and palisade tissue differentiation in Chenopodium album. Plant and Cell Physiology, 42, 1303–1310.    65 Appendix: Pearson’s product-moment correlation coefficients (r) of gm on measured traits  gm LPI 0.049 CCI 0.019 LMA -0.027 tleaf -0.108 tpal -0.006 tspo -0.317 fias 0.004 Pmes/S_P 0.036 Pmes/S_US -0.079 Pmes/S_LS -0.276 Pmes/S_T -0.084 Smes/S_P -0.267 Smes/S_US 0.041 Smes/S_LS -0.290 Smes/S_T -0.237 Na 0.052 Na/Smes/S 0.307 Sc/S -0.427 Chlo_P -0.207 Chlo_US 0.011 tcw 0.462 tgra 0.307  LPI: leaf plastochron index; CCI: chlorophyll content index; LMA: leaf mass per unit area; tleaf: leaf thickness; tpal: palisade thickness; tspo: spongy mesophyll thickness; fias: fraction of mesophyll occupied by intercellular airspace; Pmes/S_T: total surface area of mesophyll cells, per unit area of leaf surface, calculated from total perimeter of cell walls; Pmes/S_P: surface area of mesophyll cells calculated from total perimeter of cell walls in palisade tissue; Pmes /S_US: surface area of mesophyll cells calculated from total perimeter of cell walls in upper spongy mesophyll; Pmes/S_LS: surface area of mesophyll cells calculated from total perimeter of cell walls in lower spongy mesophyll; Smes/S_P: surface area of mesophyll cells exposed to airspace in palisade tissue; Smes/S_US: surface area of mesophyll cells exposed to airspace in upper spongy mesophyll; Smes/S_LS : surface area of mesophyll cells exposed to airspace in lower spongy mesophyll; Smes/S_T: total surface area of mesophyll cells exposed to airspace, per unit of leaf area Na: nitrogen per unit leaf area; Na/Smes/S: nitrogen per unit mesophyll area; Sc/S: chloroplast surface area exposed to intercellular airspace per unit of leaf area; Chlo_P: average count of visible chloroplasts per  palisade cells; Chlo_US: average count of visible chloroplasts per  upper spongy mesophyll cells; tcw: cell wall thickness; tgra: thylakoid grana thickness.  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0167666/manifest

Comment

Related Items