The control of transpiration in olive and almond: mechanisms under drought conditions

Universidad de Sevilla Facultad de Biología Departamento de Biología Vegetal y Ecología Tesis Doctoral / PhD Thesis THE CONTROL OF TRANSPIRATION IN OLIVE AND ALMOND: MECHANISMS UNDER DROUGHT CONDITIONS Dña. Celia M. Rodríguez Domínguez Directores Dr. Alfonso de Cires Segura Dr. Antonio Díaz Espejo Profesor Titular de Universidad de Sevilla Científico Titular del CSIC Dpto. Biología Vegetal y Ecología Instituto de Recursos Naturales y Agrobiología de Sevilla (IRNAS) Universidad de Sevilla Facultad de Biología Departamento de Biología Vegetal y Ecología Tesis Doctoral / PhD Thesis THE CONTROL OF TRANSPIRATION IN OLIVE AND ALMOND: MECHANISMS UNDER DROUGHT CONDITIONS Tesis Doctoral presentada por Dña. Celia M. Rodríguez Domínguez, en satisfacción de los requisitos necesarios para optar al grado de Doctora por la Universidad de Sevilla, bajo la dirección y tutorización de Directores Dr. Alfonso de Cires Segura Dr. Antonio Díaz Espejo Profesor Titular de Universidad de Sevilla Científico Titular del CSIC Dpto. Biología Vegetal y Ecología Instituto de Recursos Naturales y Agrobiología de Sevilla (IRNAS) Tutora Doctoranda Dra. Elena Fernández Boy Dña. Celia M. Rodríguez Profesora Titular de Universidad de Sevilla Domínguez Dpto. Cristalografía, Mineralogía y Química Agrícola Durante la realización de la presente Tesis Doctoral, Dña. Celia M. Rodríguez Domínguez ha disfrutado de una beca para la Formación de Personal Docente e Investigador concedida por la Consejería de Economía, Innovación y Ciencia de la Junta de Andalucía (convocatoria 2009) a Universidades públicas de Andalucía, en áreas del conocimiento consideradas deficitarias por necesidades docentes. La presente Memoria de Tesis se basa en los trabajos realizados fundamentalmente bajo el marco del proyecto “Consecuencias del control estomático de la transpiración en árboles frutales con riego deficitario ocasionado por las limitaciones impuestas por la conductividad hidráulica del sistema suelo-planta y las señales hormonales desde raíces. Desarrollo de un modelo mecanístico integrador” (AGL200911310/AGR) del Programa I+D del Ministerio de Ciencia e Innovación. NOTA: Según la Normativa vigente en la Universidad de Sevilla a 10 de abril de 2014 para optar a la Mención Internacional del Título de Doctor (Acuerdo 9.1/CG19-42012), aunque la presente Tesis será escrita en su totalidad en inglés, el resumen (p. 29) y las conclusiones (p. 205) serán también redactados en español. A mis padres Figures Fig. 2.1. Time courses of solar radiation (Rs), air vapour pressure deficit (VPD), volumetric soil water content ( θ v) in 30RDI, and irrigation amounts (IA) on three periods of the irrigation season at the experimental orchard ................................................................................................................................ 50 Fig. 2.2. Relationship between root length density (RLD) and root intersection on the trench wall per unit of surface ................................................. 52 Fig. 2.3. Spatial distribution of root length density (RLD) on a trench wall in both the Control and 30RDI treatment ...................................................... 53 Fig. 2.4. Seasonal courses of tree leaf area (AL) for the two irrigation treatments. .................................................................................................................... 54 Fig. 2.5. Time courses of air vapor pressure deficit (VPD) and transpiration estimated from sap flow measurements (Ep) .................................................. 56 Fig. 2.6. (A) Relationship between soil matric potential ( Ψ s) and soil hydraulic conductivity (Ks) for the Sanabria orchard soil at the top 0.6 m. (B) Comparison of curves of vulnerability to cavitation of two species with contrasting resistance: Olea and Vitis ................................................................ 57 Fig. 2.7. Results of plant transpiration (Ep) simulated by the SACC model for two species of contrasting vulnerability to caviation: Olea, resistant, and Vitis, sensitive ...................................................................................................... 58 Fig. 2.8. Modeling exercise with the SACC model under well-irrigated conditions. (A) Effect of increasing leaf area index (LAI) on transpiration rate (Ep). (B) Effect of increasing number of drippers on Ep, for two different LAIs ....................................................................................................................... 59 Fig. 2.9. Simulation of the evolution of canopy conductance (gc) by the BM-F model, for both Control (close circles) and 30RDI (open circles) trees vii ................................................................................................................................ 61 Fig. 2.10. Values of the BMF parameters for the four days in which we measured leaf gas exchange and leaf water potential ..................................... 62 Fig. 3.1. Time courses of soil water content (SWC) measured on the experimental dates, and photosynthetic active photon flux density (PPFD), air temperature (Ta) and leaf to air water vapor mole fraction gradient (∆w) on three experimental periods ................................................................................. 89 Fig. 3.2. Diurnal courses of (A) leaf water potential ( Ψ leaf) and (B) stomatal conductance (gs) on the three experimental dates ........................................... 90 Fig. 3.3. Evolution of gs data (points) and gs fitted with the BMF model (lines) on the three experimental dates during the study ............................... 92 Fig. 3.4. Values of the parameters obtained by adjusting the BMF model to gs data on the three experimental dates ............................................................ 94 Fig. 3.5. Leaf abscisic acid (ABA) concentrations from leaves sampled at (A) predawn and (B) mid-day on experimental dates ..................................... 95 Fig. 3.6. Relationships between predawn leaf abscisic acid (ABA) and (A) stomatal conductance, gs, and (B) the BMF model parameter that captures non-hydraulic effects, n, estimated by fitting the model ................................ 96 Fig. 3.7. Attribution of the change in stomatal conductance observed between 21 and 31 August in the water-stressed (WS) treatment to changes in BMF model parameters ...................................................................................... 97 Fig. 3.8. Hydraulic and non-hydraulic limitations to stomatal conductance ( γ h = K/(K + na∆w) and γ nh = na∆w/(K + na∆w) ........................................... 98 Fig. 4.1. Time courses of (A, B) volumetric soil water content ( θ v) measured on both well watered (WW) and water stressed (WS) treatments, (C, D) viii air temperature (Ta), air vapour pressure deficit (VPD) and (E, F, G, H) photosynthetically active photon flux density (PPFD) along the experiment .............................................................................................................................. 119 Fig. 4.2. Leaf water potential ( Ψ leaf, −MPa) and stomatal conductance (gs, mol H2O m−2 s−1) measured in leaves from different locations within the canopy for the two water treatments (WW and WS) and on the two experimental dates ............................................................................................. 120 Fig. 4.3. Leaf patch clamp pressure (LPCP) probe actual recordings during August 3 in three eastern and shaded leaves of the WS tree (lines) ........... 122 Fig. 4.4. Relationships between the normalized output leaf patch pressure (P’p, %) and the leaf water potential ( Ψ leaf, −MPa) measured on the two experimental dates in different leaf locations within the canopy and for the two water treatments (WW and WS) .............................................................. 123 Fig. 4.5. Evolution of gs data and gs fitted with the BMF model on the two experimental dates in different leaf locations within the canopy and for the two water treatments (WW and WS) .............................................................. 126 Fig. 4.6. Values of the parameters obtained by fitting the BMF model to gs data on the two experimental dates in different leaf locations within the canopy and for the two water treatments (WW and WS) ........................... 127 Fig. 4.7. Diurnal variation of soil-to-leaf hydraulic conductance (Kvar) obtained by exactly fitting the BMF model to gs data on the two experimental dates in different leaf locations within the canopy and for the two water treatments (WW and WS) ................................................................................ 128 Fig. 4.8. Diurnal variation of soil-to-leaf hydraulic conductance (Kvar) plotted against leaf water potential ( Ψ leaf) measured in eastern and shaded leaves and mixing the two water treatments and the two experimental dates .............................................................................................................................. 129 ix Fig. 4.9. Relationships between the normalized output leaf patch pressure (P’p, %) and the absolute leaf turgor pressure modeled with the BMF model (Pmodel) on the two experimental dates in different leaf locations within the canopy and for the two water treatments (WW and WS) ........................... 130 Fig. 5.1. Leaf patch clamp pressure probe measurements on a 2-m tall olive tree subjected to irrigation ⁄ non-irrigation cycles under well-defined laboratory conditions ....................................................................................... 145 Fig. 5.2. Pressure transfer through the leaf patch under irrigation and nonirrigation conditions as predicted from Equation (5.1) ............................... 146 Fig. 5.3. Effects of non-irrigation ⁄ irrigation cycles on the time constant τ of the exponential Pp decrease measured in state I after switch off of the light, as well as the exponential Pp decrease after switch on of the light and the exponential Pp increase after switch off of the light measured in state III (for approximation of the Pp curves in state III by exponentials, see Fig. 5.5) ............................................................................................................................. 147 Fig. 5.4. Calibration of the leaf patch clamp pressure, Pp, measured in state I through short-term measurements of cell turgor pressure, Pc ................... 149 Fig. 5.5. Part of a long-term measurement of diurnal changes of Pp measured on east-oriented leaves of a control tree (A) and a 60RDI tree (B; RDI = regulated deficit irrigation) under field conditions ....................................... 150 Fig. 5.6. Typical images of cross-sections of olive leaves under well-watered conditions (A; state I) and severe water stress (B; state III) .......................... 151 Fig. 5.7. Calculations of Fa,Pc≈0 from Pp changes measured in state III ..... 155 Fig. 6.1. Seasonal changes of (A) the potential (ETo) and crop (ETc) evapotranspiration, (B) the collected precipitation (P) and the irrigation amounts (IA) supplied during each irrigation treatment, and (C) the relative extractable water (REW) for each treatment ................................................ 172 x Fig. 6.2. Average values (n = 8) of (A) predawn leaf water potential ( Ψ pd) and (B) midday stem water potential ( Ψ stem) measured on representative trees of each treatment during the experimental period .............................. 173 Fig. 6.3. Part of a long-term measurement of diurnal changes of the patch pressure Pp (A) and of the sap flow rate Q (B) ............................................. 174 Fig. 6.4. Plots of diurnal changes of the Q values versus the corresponding diurnal changes of the Pp values together with vapor pressure deficit of the air (Da) and solar radiation (Rs) values recorded on the given day and the day before. Typical examples are given for Q peaking preceding Pp peaking (A) and for Pp peaking heading Q peaking (B and C) .................................. 176 Fig. 6.5. Part of a long-term measurement of diurnal changes of the patch pressure Pp (A) and of the sap flow rate Q (B). The corresponding changes of solar radiation (Rs) and vapor pressure deficit of the air (Da) are given (C). Pp was measured on east-oriented leaves of a 30RDI tree .................. 178 Fig. 6.6. Plot of the changes of daily maximum sap flow rates (Q) measured between DOY 168 (turgescent Pp state I) and DOY 209 (inversed Pp state III). Data were taken on a 30RDI tree ........................................................... 180 Fig. 6.7. Plots of diurnal - changes of the Q values versus the corresponding changes of the Pp values recorded on the 30RDI tree in Fig. 6.3 when the Pp curves became inverted (state III). Also shown are the vapor pressure deficit of the air (Da) and solar radiation (Rs) values measured on the given day and the day before. Typical examples are given for Q peaking preceding the Pp minimum value (A) and for Pp minimum value heading Q peaking (B) .............................................................................................................................. 181 Fig. 6.8. Concomitant measurements of diurnal changes of Pp measured by a LPCP probe and of the leaf water potential Ψ leaf (mean ± SE, n = 8) measured with a pressure chamber on representative 30RDI (A) and 60RDI (C) as well as on Control trees (B and D) ...................................................... 182 xi Fig. 6.9. Linear relationships between the output pressure Pp values and the leaf water potential Ψ leaf values measured on Control (A) and 30RDI (B) trees on the same days as in Fig. 6.6 .............................................................. 183 Extra Fig. 2.1. Agreement between soil matric potential ( Ψ s) and pre-dawn leaf water potential ( Ψ pd) along the season ................................................... 217 Extra Fig. 3.1. Measurements of stomatal conductance in relation to PPFD, validating the assumption that the response of gs to irradiance is homogeneous in almond under field conditions ......................................... 223 xii Tables Table 2.1. Soil textural and physical properties obtained by the Rosetta software ...................................................................................................................... 51 Table 2.2. Seasonal evolution of main physiological variables in both Control and regulated deficit irrigation trees (30RDI) .......................................... 55 Table 3.1. Maximum carboxylation rate (Vc,max), maximum potential electron transport rate (Jmax) and mesophyll conductance (gm) measured at the beginning and end of the experiment ............................................................... 83 Table 4.1. Maximum carboxylation rate (Vc,max), maximum rate of electron transport (Jmax) and mesophyll conductance (gm) measured along the experiment .......................................................................................................... 114 Table 4.2. Osmotic pressures measured at dawn ( π d) on the two daily cycle measurements .................................................................................................... 125 Extra Table 3.1. Parameter values used in this chapter for responses of photosynthetic parameters to temperature .................................................... 226 xiii xv Most used abbreviations and symbols Symbol or Abbreviation Description Units ABA Abscisic acid - A L Leaf area m −2 tree −1 A N Net CO 2 assimilation rate  mol m−2 s−1 A R Root area m 2 tree −1 c a Ambient CO 2 concentration  mol mol−1 c i Intercellular CO 2 concentration  mol mol−1 D a Air vapor pressure deficit kPa DI Deficit irrigation - DOY Day of year - E Leaf transpiration rate mmol H 2 O m−2 s−1 E p Plant transpiration per leaf area mmol m −2 s −1 ET c Crop evapotranspiration mm ET o Potential evapotranspiration mm F a Attenuation factor of the leaf transfer function - g c Canopy conductance mol m −2 s −1 GCM General circulation model - g m Mesophyll conductance to CO 2 mol m −2 s −1  g s Stomatal conductance mol m −2 s −1 HPV Heat - pulse velocity - IA Irrigation amounts L tree −1 day −1 IN Irrigation needs mm J max M aximum potential electron transport rate  mol m−2 s−1 K L eaf - specific hydraulic conductance mmol m −2 s −1 MPa−1 K c Crop coefficient - K r Reduction coefficient related to the percentage of ground covered by the crop - LAI Leaf area index m 2 leaf area m−2 ground Introduction 4 Stomata are minute pores mainly located in the lower epidermis of the leaves (abaxial surface) and bounded by a pair of specialized guard cells. These guard cells play a vital role by regulating the opening and closing of the stomata through increasing or declining, respectively, their turgor pressures. This dynamic gas exchange between the leaves and the atmosphere represents a key process for CO2 and water fluxes, and is strongly linked with plant productivity since stomata control transpiration and photosynthetic CO2 uptake (Jones & Tardieu, 1998). Environmental factors such as light intensity and quality, temperature, water vapor concentration in the air, leaf water status, and intracellular CO2 concentrations are sensed by guard cells, integrating all of these signals into well-defined stomatal responses. Water uptake by the roots is carried through the vascular system of the plant (xylem) to the cell walls of the mesophyll, where it evaporates into the sub-stomatal cavity. Water vapor is released by diffusion through the stomatal pores in the process called transpiration. Despite some cuticular transpiration exists, most leaf transpiration occurs through the pores of stomata. Only a small percentage (~2 %) of the water absorbed by roots is used by the plant to supply growth or to play in the biochemical machinery for photosynthesis or other metabolic processes (Taiz & Zeiger, 2010). When the water is evaporated, a large tension (negative hydrostatic pressure) is developed, pulling water through the xylem. This mechanism, established over 100 years ago as the ‘cohesion-tension theory of sap ascent’, explains water movements through plants relying on basic physical properties of water and on some assumptions: the high cohesive forces of water are necessary to sustain large tensions in the xylem water columns; when a critical tension is reached in the xylem vessels, the soil-plant-atmosphere-continuum (SPAC) that water forms is broken by the process called ‘cavitation’, resulting in gas-filled, non-functional vessels or ‘embolism’; transpiration lowers the leaf water potentials, causing water to move up; and, finally, the sun provides the energy needed for that movement, by increasing the temperature of both leaf and the air surrounding, and driving the evaporation of water. But, this Chapter 1 5 theory has been recently criticized, generating still a fervent debate (Steudle, 2001; Angeles et al., 2004; Zimmermann et al., 2004). Stomatal conductance (gs) is the conductance associated with diffusion of gases (CO2 and water vapor) through the stomatal pore. Four hundred water molecules are lost from leaves per molecule of CO2 fixed by photosynthesis (Taiz & Zeiger, 2010). Thus, when soil water availability is scarce, a trade-off between avoiding dehydration and allowing sufficient CO2 fixation for photosynthesis is raised. As a consequence of the stomatal control of transpiration, photosynthetic limitations appear. Traditionally analyzed in terms of ‘stomatal’ and ‘non-stomatal’ limitations (Flexas & Medrano, 2002), at the present, photosynthetic limitations under drought conditions are composed of ‘difusional’, consisted of gs and mesophyll conductance (gm) to CO2, and ‘non-difusional’ or ‘biochemical’ limitations, integrated in metabolic changes of the ribulose-1,5-bisphosphate carboxylase/oxygenase (Rubisco) related to the photosynthetic capacity of the plant (Flexas et al., 2013). A quantitative estimation of the relative significance of gs, gm and the maximum capacity for carboxylation (Vc,max) in limiting net photosynthesis (AN) has been proposed by Grassi & Magnani (2005). The control of transpiration by stomata in both wild and cultivated plants has been proved in many studies (Irvine et al., 1998; Franks, 2004), including olive (Fernández et al., 1997; Moriana et al., 2002; Testi et al., 2006) and almond species (Klein et al., 2001; Romero & Botía, 2006). In response to environmental changes, gs is controlled by plant and soil hydraulics being down-regulated to prevent runaway xylem embolisms (Sperry et al., 2002). Furthermore, water deficit strongly stimulates abscisic acid (ABA) biosynthesis by roots and leaves and influences gs mediating stomatal closure (Davies & Zhang, 1991; Sauter et al., 2001; Holbrook et al., 2002; Bauer et al., 2013). Because stomata are the gatekeepers of terrestrial photosynthetic gas exchange, they become a crucial component in all of the Introduction 6 above processes and understanding their dynamic behavior is imperative to comprehend the control of transpiration by plants. The control of transpiration in precision agriculture Understanding the basis of the control of transpiration is crucial in precision agriculture within a context of global changing conditions and restrictive water availability. The Mediterranean basin is recognized as one of the major areas where agriculture was originated, with the first domestication events occurring in the Fertile Crescent about 12 000 years ago (Zeder, 2008), followed by the spread of agricultural practices along the European and African Mediterranean shores. This expansion of agriculture was the consequence of two main advantages over the nomadic lifestyle of hunter-gatherers (Smith, 2001). First, the increment of food yield per fertile land area allowed maintaining a greater human population. And second, these novel food producing societies were sedentary and could store the accumulated surplus, resulting in the development of new technologies and stratified societies evolving to what we know at the present-day. Since these early events of plant domestication, humans have developed different methods to hold up and enhance those advantages. Among these new techniques, the emergence of irrigation to avoid water deficits that reduced production was crucial for a constant provision to a continuously larger human population. The current increase of cultivated areas promoted by the expansion in human population and thus, a higher demand for food, has led irrigated agriculture to be the largest consumer of water in the world, becoming up to 80 % in arid and semiarid environments (Fereres & Soriano, 2007). This causes a diminution of water availability for other sectors of the society, which is exacerbated with the effects of the climate change, further accentuated in areas with Mediterranean climate. More frequent and severe drought combined with high temperatures has been recognized as a potential impact of global warming on agriculture. Rainfall Chapter 1 7 will become more unpredictable and extreme weather events, such as heat waves and floods, will occur more frequently (IPCC, 2013). In this context, implementation of more efficient irrigation strategies is compulsory to reduce agricultural water use, to make water resources more sustainable and, at the same time, to meet the food needs of a growing population. Furthermore, the idea of increasing crop yield per unit of cultivated area has turned into maximizing ‘water productivity’ or ‘water use efficiency’, i.e. amount of yield produced per unit of water used. In order to improve water productivity in agriculture, continuous search of new precision irrigation techniques and strategies has motivated during the last decades several and insightful studies (Fereres et al., 2003; Morison et al., 2008; Jacobsen et al., 2012). The use of drip irrigation techniques has the main advantages over other techniques of water savings, due to improvements in water circulation, increasing fertilizer use efficiency and decreasing soil salinity. Moreover, this irrigation technique coupled with applying deficit irrigation (DI) seems to be the most suitable strategy for a rational use of water in fruit tree orchards in arid and semiarid areas (Fereres et al., 2003). DI is defined as the amount of water supplied to the orchard below its water needs. An optimal management of DI can allow considerable water savings without causing negative impacts in production and even, at times, improving quality. However, mismanagement of DI strategies may cause water deficits in periods when plants are more sensitive to water stress, reducing both production and productive life span of the orchard (Fereres & Evans, 2006). Thus, the application of regulated deficit irrigation (RDI) that Chalmers et al. (1981) and Mitchell & Chalmers (1982) started to use in peach orchards proved that it was possible to reduce the water use without penalizing production. RDI is defined as the reduction or withdraw of water supply in plant phenological periods in which the orchard is more tolerant to water deficits, imposing a controlled water stress. This irrigation strategy seems to be the most advisable tool to Introduction 8 improve water use efficiency and orchard yield in semiarid conditions (Boyer, 1996). Over the last two decades, the RDI strategy has been extensively investigated, still increasing, in the main fruit tree crops, such as citrus (González-Altozano & Castel, 2003; Pérez-Pérez et al., 2008), almond (Romero et al., 2004; Girona et al., 2005; Goldhamer et al., 2006), olive (Giorio et al., 1999; Iniesta et al., 2009; Fernández et al., 2013), peach (Dichio et al., 2007), vine (McCarthy, 1997; Cifre et al., 2005; Santesteban et al., 2011), or plum trees (Intrigliolo & Castel, 2010). But, knowing when, how much and how to apply this irrigation strategy is still a pending and challenging task not only for orchard managers, but especially for the scientific community. It depends on the species, and even on the variety within a species, and on the soil and weather conditions of the orchard location. In the present Thesis, we will focus on two high resistance species to drought typical from Mediterranean climate: olive and almond. Physiological and agronomical features of two Mediterranean species under drought: olive and almond Drought and water stress Water is one of the main abiotic factors, together with light, temperature and mineral elements in the soil, influencing and constraining plant physiological and biochemical behavior (Taiz & Zeiger, 2010). Land plants have been coping with drought and water stress since they first left the seas and colonized dry land. ‘Water deficit’ (insufficient water availability) occurs in most natural and agricultural environments and is mainly caused, but not only, by short to long periods without precipitations. ‘Drought’ can be defined as the meteorological period in which partial or none precipitation events happen, resulting in soil dehydration, diminution of soil water Chapter 1 9 reservoirs and, hence, in plant water deficit. Indeed, this environmental dynamic state is considered to be the most frequent cause of water stress in plants (Boyer, 1982). ‘Water stress’ is a functional and structural plant response to low water availability condition. This situation of water deficit is not only provoked by rainfall scarcity, but also, and overall in Mediterranean climate, by the combination of high temperatures, light intensities and atmospheric demand. All of these environmental factors cause a large variety of effects in soil and plants, altering most of their physiological processes. After a heavy rainfall or an abundant irrigation event and once excess water has been allowed to drain away, the water content of a soil reaches ‘field capacity’. As soil dries down, its hydraulic conductivity decreases and the ‘permanent wilting point’ can eventually be reached, i.e. the soil water content at which plants cannot regain turgor upon rehydration (usually at about −1.5 MPa, but it depends on the species). At this point, water delivery to the roots is too slow to allow the overnight rehydration of plants that have wilted during the day (Taiz & Zeiger, 2010). The control of irrigation occurs between field capacity and wilting point, and it is in this range of available water where most of the plant regulation and physiological response to water stress occurs. Normally, the first symptoms of water stress in plants are cellular dehydration, inhibition of cell expansion and reduction in cell turgor, that result in reduction of stomatal opening to prevent desiccation (Flexas & Medrano, 2002). Stomatal closure in response to dehydration is almost always an active, energy-dependent process rather than a passive one (Buckley, 2005). Abscisic acid (ABA) mediates the solute loss from guard cells that is triggered by a decrease in the water content of the leaf and induces stomatal closure under water stress conditions. Subsequently, photosynthesis is unavoidably reduced due to decreased CO2 availability at chloroplast level. When water stress becomes more severe, photosynthetic capacity is reduced (through decreasing carboxylation efficiency), xylem dysfunctions are induced, osmotic adjustment (net accumulation of solutes to lower water potential during drought periods) appear or disruption of phloem function occur. Introduction 10 An understanding of the impact, mechanisms and traits underlying drought tolerance is essential to improve water productivity. But first, it is necessary to understand some concepts. Plant species differ widely in their capacity to cope with drought. Response mechanisms to drought are mainly divided in ‘avoidance’ and ‘resistance’ behaviors that profoundly depend on the plant capacity to maintain water status (Kozlowski & Pallardy, 2002; Tyree et al., 2002). Within the mechanisms of drought resistance two types of responses can be distinguished: ‘desiccation avoidance’ and ‘desiccation tolerance’. Desiccation-avoiding responses (through increasing access to water and reducing water loss) include leaf shedding and changes in leaf orientations, early stomatal closure, sunken stomata, low cuticular conductance (through abundant leaf waxes), enhanced water storage in plant sinks, strong development of palisade mesophyll, low resistance to water flow in vascular tissues, deep roots and extensive root growth. Desiccation-tolerance responses (through physiological changes that allow continued water transport, gas exchange and cell survival at low water content) usually involve osmotic adjustment, changes in elasticity of tissues, decreased vulnerability of xylem to embolism and molecular-level changes. Olive and almond are considered to be drought resistant species, but the mechanisms to cope with water stress are different between them. Olive Olive (Olea europaea L.; Oleaceae) is a common crop of the Mediterranean basin and native to the coastal areas of the eastern Mediterranean. Olive on crop probably started 6000 years ago, when the first olive growers began to select from the surrounding wild olive forests significantly better individuals due to their productivity, fruit size and oiliness, and adaptation to the environment. Vegetative propagation has maintained the characteristics of those initially selected cultivars that constituted the first varieties (Barranco, 2001). The Mediterranean climatic region is characterized by long periods of drought during the dry season (from May to September) when temperature Chapter 1 11 and atmospheric demand are high, and a mild wet season from October to April. At present, of the estimated total of 9.5 million ha of olive orchards in the world (producing 20.8 million t), two-thirds (6 million ha) are extended through the Mediterranean region. Spain has the largest area (2.09 million ha), and is also the largest producer of table olives and olive oil (total 8.01 million t), comprising olive oil over 90 % of the production (FAO, 2012). Its capacity to grow and produce acceptable yields under severe drought conditions, and the increased demand for olive products, especially olive oil, driven by health-related benefits (Hu, 2003), have led in the last decades to an expansion of olive orchards and to an increasing of studies on olive biology and growing (Fernández, 2014b). Although olive is considered to be drought-resistant, it responds well to irrigation, which in fact can improve yield (Iniesta et al., 2009). A correct equilibrium between water scarcity in arid and semi-arid areas producing olives, and sustainable irrigation strategies is necessary to maintain good production and quality (Carr, 2013). For that purpose, it is essential to disentangle the physiological mechanisms of olive tree to respond and adapt to water stress and how to use this knowledge to improve crop management practices (Fernández & Moreno, 1999). According to Rapopport (2001), the olive tree under cultivation can attain a height of 4-5 m maintained by pruning. Canopy is normally rounded, almost lobed, and their branches tend to make it fairly dense. Pruning practices are designed to allow light to penetrate into the canopy and to promote fruiting sites. The canopy shape is of considerable importance since it can modify CO2 assimilation rate and water use efficiency (Diaz-Espejo et al., 2002) and, hence, it can affect vegetative and reproductive developments. Olive is considered a xerophytic evergreen and sclerophyllous tree species and therefore, it has morphological and physiological characteristics that confer it high capacity to resist water stress (Fernández, 2014b). The growth and reproductive olive biennial cycle contributes to alternate bearing, with years of intense fruiting tending to be Introduction 12 followed by years of restricted flowering and reduced crop load. Leaf upper, adaxial surface is covered by a waxy cuticle. Palisade parenchyma usually consists of two-to-three highly packed layers of elongated cells, and spongy mesophyll anatomy greatly depends on leaf water status (Ehrenberger et al., 2012b). In the lower, abaxial surface of the leaf, stomata are covered by a dense network of trichomes, providing a very effective control of transpiration (Connor, 2005). Under dry conditions stem develops a thick cork layer covering the living bark tissues, thus protecting against sunburning. Below the bark there is the phloem, the cambium, and the xylem. Narrow xylem vessels with low hydraulic conductivity reduce risk of embolism and allow olive to withstand water potentials below turgor loss point with minor xylem embolisms (Torres-Ruiz et al., 2013a). This low vulnerability to embolisms is different depending on the plant organ. Thus, hydraulic segmentation (Zimmermann, 1983; Tyree et al., 1993) occurs in olive (Torres-Ruiz et al., 2013b), as in other Mediterranean species (Martínez-Vilalta et al., 2002), making leaves effective in reducing wholeplant transpiration and, hence, in avoiding the spread of embolism in other plant organs. The olive root system is adapted to water scarcity, and its depth, shape and lateral extension will be dependent on the plant variety and on the type, aeration and water content of the soil (Fernández et al., 1991). For olive trees with localized irrigation, the greatest root length densities of fine (Ø < 0.5 mm), active roots, are found in wetted soil volumes close to the drippers, with a favorable balance between air and water for root growing. In dual soils, characterized by a sandy top layer and a clayey bottom layer of high resistance to penetration, roots may only explore the top layer, and penetrate deeper layers due to soil cracks and favorable aeration when soil dries (Diaz-Espejo et al., 2012). Water supply, tree age, plant density and soil characteristics greatly affect the root/canopy ratio, being usually larger in rain-fed than in irrigated olive trees (Fernández et al., 1991). Furthermore, the well-known capacity of olive to take up water from drying soils may be allowed by physiological mechanisms aiming to maintain turgor pressure, such as osmotic and elastic adjustments Chapter 1 13 (Fernández, 2014b). Stomatal closure mediated by chemical and hydraulic signals is still a matter of debate (Schachtman & Goodger, 2008; Pantin et al., 2012; Christmann et al., 2013; Dodd, 2013; Franks, 2013). Droughtinduced ABA from roots and long-distance signaling are dominated by complex mechanisms (Davies & Zhang, 1991). Some authors argue that stomata mostly respond to ABA generated in the leaf, rather than in the roots (Bauer et al., 2013). Recently the response of stomata to soil drying seems to depend mostly on hydraulic signals rather than on chemical signals in olive, especially under saturating light and high evaporative demand (e.g., at midday, Diaz-Espejo et al., 2012). Therefore, a good election of the irrigation techniques and strategies has important effects on how olive trees behave in arid and semi-arid conditions, where water for irrigation is scarce. During the last years, new olive growing systems designed for improving short-term crop performance have been developed and gradually implemented. Among them, hedgerow olive orchards with high plant densities, also called super-high-density orchards (plant densities close to 2000 trees ha−1), have been gaining importance, occupying nowadays more than 40 000 ha worldwide (Gómez-del-Campo, 2013). Besides early yield after planting, these orchards are designed for high yield when the crop is established, compact bearing, self-fertility, limited vigor and mechanization, especially during harvesting (Rius & Lacarte, 2010). Among the large list of olive tree cultivars, cv. ‘Arbequina’ is well adapted to these far-extended commercial orchards and it has been the cultivar in which our studies have been focused during the last four years. Olive responds well to deficit irrigation, since it is well-adapted to stressed conditions and has a marked productive response to favorable water supplies. In Mediterranean regions, potential evapotranspiration (ETo) range from 1000 to 1400 mm year−1, and crop evapotranspiration (ETc), from 560 to 800 mm year−1 depending on environmental conditions, crop characteristics, and orchard management (Fernández & Moreno, 1999). Considering that mean annual precipitations values are lower than 600 mm and irrigation needs in super-high-density Introduction 20 ultimate effect on production. Although these methods have several advantages and have been widely used in research, there are still a number of practical difficulties for implementing their use in commercial orchards (for a review on this topic, see Fernández, 2014a). Conventional plant-based methods to monitor water stress include leaf and stem water potentials (Naor et al., 2006), stomatal conductance, shoot length growth and leaf area. Even though there was often homeostasis of leaf water potential between different soil moisture regimes, rapid temporal fluctuations are often observed as a function of environmental conditions (such as passing clouds). This makes the interpretation of leaf water potential as an indicator of irrigation-need doubly unsatisfactory (Jones, 1990). As a partial solution for that, stem water potential at midday (measured on leaves enclosed in darkened plastic bags for some time before measurement and allowed to equilibrate with the xylem water potential) has been proposed as a more robust indicator of water status (Shackel et al., 2000). Furthermore, pre-dawn leaf water potential has been used as an estimator of the soil water potential. All water potentials are measured with a Scholander-type chamber (Scholander et al., 1965) and hence, are highly invasive, destructive and time consuming. Less invasive methods are stomatal conductance, shoot growth and leaf area measurements. Nevertheless, they are still largely time and labor consuming. None of these traditional plant-based methods are suitable for automatic and continuous measurements. Thus, in the last decades, new water stress monitoring methods have been developed for non-destructive, automatic and continuous data recording, easily implemented with data transmission systems for a nearly real time access to the collected records from a remote computer. Most of them are highly sensitive and capable of working under field conditions for long periods of time, meeting most of the criteria for both reliable monitoring of water stress and irrigation scheduling (Jones, 2004; Fernández, 2014a). Among these methods, those based on measurements or estimations of sap flow, trunk diameter variation and leaf Chapter 1 21 turgor pressure are the most widely studied. Sap flow (SF) methods have a potential for in situ determinations of plant water consumption and transpiration dynamics (www.wgsapflow.com). Because SF rates are expected to be sensitive to water deficits and especially to stomatal closure, the use of SF measurements for irrigation scheduling has been tested in a diverse range of crops, including comparisons with other water stress indicators (Fernández et al., 2001, 2008b; Intrigliolo & Castel, 2006; Ortuño et al., 2006; Conejero et al., 2007). Although changes in transpiration rate estimated by sap flow are largely determined by changes in stomatal aperture, transpiration is also influenced by other environmental conditions such as atmospheric water demand, and reference trees are often used to cancel-out the effect of environmental variables (Fernández et al., 2008b, 2011a). The usefulness of trunk diameter variation (TDV) records both for monitoring water stress and scheduling irrigation has been evaluated for a great number of species (Intrigliolo & Castel, 2006b; Fernández & Cuevas, 2010; Ortuño et al., 2010), and comparative studies between TDV and other water stress indicators have also been made (Ehrenberger et al., 2012a; Cuevas et al., 2013). TDV devices are relatively easy to install, their outputs are quite sensitive to water stress and can be continuously and automatically recorded. However, the TDV outputs are affected by seasonal growth patterns, crop load, plant age and size, and other factors besides water stress (Fernández & Cuevas, 2010). Thus, continuous readjustments are needed along the irrigation season. Furthermore, detailed interpretations of TDV signals require strong knowledge on plant physiology to disentangle the role of each trait on the trunk variation. The leaf patch clamp pressure probe, or ZIM probe (Zimmerman et al., 2008) is a relatively novel plant-based sensor able to record automatically and continuously the so-called leaf patch output pressure (Pp), which is inversely correlated with the leaf turgor pressure (Pc), when Pc > ca. 50 kPa (Rüger et al., 2010; Ehrenberger et al., 2012a, 2012b), a variable closely related to leaf water potential and stomatal conductance (Ache et al., 2010). It has been used in several plant species and crops to precisely monitor water stress (Westhoff et al., 2009; Rüger et al., Introduction 22 2010; Fernández et al., 2011b; Lee et al., 2012; Bramley et al., 2013). Furthermore, this variable is one of the physiological variables recognized to be among the most sensitive to water stress (Jones, 2004, 2007), although it still requires intense evaluation under field conditions to prove its applicability. The present Thesis dedicates three Chapters (4, 5 and 6) to the study of this plant-based sensor. Moreover, Chapter 4 combines these probe signals with a process-based stomatal conductance model (see next section). Modeling leaf gas exchange: the quest for the right model Once recognized the crucial role of stomatal functioning in the control of transpiration and fluxes of CO2 between the leaf and the atmosphere, modeling stomatal conductance (gs) would be one of the most effective and valuable tools for improving our understanding of the regulation of stomatal conductance to changing environmental conditions and to water stress, as well as for integration, simulation and prediction purposes (Damour et al., 2010). Indeed, a trustworthy modeling of gs is also required for successful modeling of plant transpiration, necessary for designing more efficient and water-saving cropping systems. This is especially suitable in fruit trees where a low decoupling factor exists (Jarvis & McNaughton, 1986). Under these conditions, there is a high coupling of leaves and canopies to atmospheric conditions and, as a consequence, a high importance of the aerodynamic conductance in controlling water flux from vegetation to atmosphere decreases. Thus, stomata strongly control transpiration rates. The boundary layer reduces this effect, i.e. as the ratio of boundary layer conductance to stomatal conductance decreases (or, in other words, the boundary layer resistance increases comparing to stomatal resistance), the decoupling factor increases. The coupling to the atmosphere of fruit tree canopies is modulated daily from wind velocity, incoming radiation and evaporative demand, and seasonally from a high influence of Chapter 1 23 the water deficit endured by trees (Aranda et al., 2012). This translates in the latter that transpiration becomes highly dependent on the sensitivity of stomata to changes in water vapour deficit of atmosphere. Although global models are very useful to accurately predict and assess impacts of climate change on carbon and water cycles, they are generally incapable of disentangling the mechanisms through which stomata respond to environmental conditions, especially water stress, which has become the real Achilles´ heel in gs modeling. Therefore, including accurate predictions of plant gas exchange with stomatal conductance models at the leaf level are needed to implement these models (Egea et al., 2011b; Buckley & Mott, 2013). The majority of stomatal modeling approaches at leaf level are empirical (data-based), based on statistical correlations between environmental or internal factors and gs, or semi-empirical, built on physiological hypotheses, but still combined with empirical functions, and very few are really mechanistic (process-based). Nevertheless, both empirical and mechanistic approaches are difficult to completely separate, since even the most empirical models are at least partially mechanistic and even the most detailed mechanistic models must resort to empiricism at some scale (Buckley & Mott, 2013). There is also another approach for modeling gs: economicor optimization-based. This approach is focused on why stomata behave as they do leading into the optimization of carbon gain versus water loss, and how that behavior impacts other aspects of plant form and function (the so-called stomatal optimization theory) (Cowan & Farquhar, 1977). Responses of stomata to environmental factors have been extensively described (Jarvis, 1976; Jones, 1992; Monteith, 1995). These responses are mediated from short to long distances by many internal signals such as hormones, reactive oxygen species (ROS), CO2 concentration in the leaf intercellular space and hydraulic signals (Hetherington & Woodward, 2003). However, the links between environmental conditions, how they are sensed and translated into the whole plant functioning and how they end up into Introduction 24 stomatal responses, are still poorly elucidated. Disentangling these mechanisms and relationships will help to develop more integrative approaches of stomatal responses to environmental influences where possible couplings and interactions could be taken into account. A particularly powerful approach might be to parameterize plant and leaf models of hydraulic function, such as the SACC model of Sperry et al. (1998, 2002), concurrently with a stomatal model. However, the main challenge remains incorporating the effect of water stress in the models, gaining in the capacity to simulate plant functioning under limited water supply conditions (Damour et al., 2010; Egea et al., 2011b). This section reviews the most used gs models, mainly focusing on empirical and mechanistic approaches capable to account for multiple environmental influences with special attention to water stress conditions. The stomatal conductance model or combinations of models selected will depend on the final intended use and objectives of the research. Empirical models are often numerically simpler than mechanistic models, and they are more recommended for those users who want to combine them within a larger model of global or canopy level processes. More mechanistic models are often more mathematically complex, but the physical and biochemical basis in which they fall on allow their use for investigating the cellular and sub-cellular processes involved in environmental sensing, signal transduction, and ion movements (Buckley & Mott, 2013). Among the empirical models, the ‘Jarvis’ model (Jarvis, 1976) is a simplistic modeling approach based on the observed responses of gs to environmental factors. This multiplicative model of environmental influences integrates responses to irradiance (PPFD), leaf temperature (Tl), air vapor pressure deficit (VPD), CO2 concentration in the air (ca) and leaf water potential (  leaf), assuming that each response is independent of the others. Following modifications expressed these responses as reduction factors of a maximal stomatal conductance (gs,max) related to stomatal closure or limiting gs: Chapter 1 25 (1.2) )()()VPD()()PPFD( max, leafalss fcffTffgg     1,0  f. Although these limiting factor-based approaches can explain 95 % of the observed variation of gs, interactive effects between environmental factors are not taken into account and their empiricisms make necessary new parameterization for each new environmental condition. However, they have been successfully tested in the field (e.g. olive or walnut trees) linking with the biochemical model of photosynthesis of Farquhar et al. (1980) and incorporating the effect of soil water deficit (Le Roux et al., 1999; DiazEspejo et al., 2006). Furthermore, its modular structure makes it easy to include into larger models, such as general circulation models (GCM) (Egea et al., 2011b), and explains its still wide use by the scientific community. One of the most widely used empirical models is the ‘BWB’ model (Ball et al., 1987) and variations thereof. This coupled photosynthesis-stomatal conductance model is basically based on the correlation usually found between gs and AN. It is able to describe the stomatal response to light, humidity and CO2, being gs described as a function of net photosynthesis (AN), CO2 concentration at the surface of the leaf (cs), the relative humidity at the surface of the leaf (hs) and the residual stomatal conductance when AN is zero (g0): (1.3) s sN schA mgg  0, where m is an empirical parameter that varies among leaves and among different water stress conditions. As it can be observed from the Equation (1.3), m is the slope of the commonly linear relationship between gs and AN. It is also widely used in canopy models and GCMs (Egea et al., 2011b). Since stomata do not respond directly to cs per se, but rather to CO2 concentration in the intercellular spaces (ci) (Mott, 1988), for predicting purposes, the so-called ‘supply function of CO2 diffusion rate’ must be coupled to a model of AN (Farquhar et al., 1980): Introduction 26 (1.4)          s i s N 1 6.1 c c c A gs. Leuning (1990, 1995) proposed modified versions of the ‘BWB’ model to enhance its physiological meaning. Thus, CO2 compensation point (  *) is taken into account, subtracting from cs, to prevent AN from becoming negative at low cs, which could lead to negative gs values. Later, the relative humidity term (hs) was replaced with an inverse hyperbolic response to the leaf-to-air vapour pressure deficit (VPD): 1/(1 + VPD/D0), where D0 is an empirical parameter. Indeed, Mott & Parkhurst (1991) demonstrated experimentally that stomatal responses to humidity are really responses to water-loss rates (transpiration), and thus showed that stomata sense leaf transpiration rate (E) rather than air humidity. Furthermore, the hyperbolic relationship between gs and VPD also arises in mechanistic models from the effect of transpiration on turgor pressures of cells in the stomatal complex (e.g. Dewar, 2002; Gao et al., 2002; Buckley et al., 2003). The revised versions can be summarized as: (1.5)            0 s N 0s VPD 1* D c A mgg  . ‘BWB’ and ‘Leuning’ models are still extensively used. This is mainly because they both represent a good compromise between ease-to-use, explicative power and predictive accuracy in various experimental conditions. Although these models have been implemented to simulate gs in water stress conditions (e.g. introducing a function of soil water content or a function of ABA or  leaf), they are still rather empirical and hence, based on observed relationships which lead to complicated simulations in a wide range of environmental conditions. Chapter 1 27 With a better understanding of plant physiological mechanisms involved in the response to water stress and as a powerful tool to improve it, process-based models have been proposed (Buckley & Mott, 2013). Some of them are based on the effects of leaf water potentials and xylem ABA concentrations on gs (Gutschick & Simonneau, 2002). Others use hydraulic models, where water flux through the whole tree equals transpiration, to simulate gs. In these approaches, the hydraulic tree architecture is represented by different organs, from the roots to the leaves, and specific hydraulic conductivities for each organ are estimated (Sperry et al., 1998). But none of these mentioned mechanistic models consider at the same time the whole combination of stimuli that influence gs: hydraulic and nonhydraulic factors. ‘Hydromechanical’ models aim to integrate gs as a function of water balance and turgor regulation of guard cells. This approach has led to the emergence of common issues that account for the influences of guard cell and epidermal turgor pressures (Pg and Pe) on gs. Both Pg and Pe are related to water potential (  ) and osmotic pressure (  ), by the standard expression of plant-water relations: Pg =  g +  g, and Pe =  e +  e (taking the convention that osmotic pressure is positive). Some assume that the differences between both Pg and Pe govern stomatal movements, and that stomatal response to water loss is mediated by a feedback mechanism (the so-called ‘hydropassive feedback’) where transpiration causes diminution of water potential and reduces Pg, leading to stomatal closure (Dewar, 2002). On the other hand, for other hydromechanical models, stomata respond simply through the direct effect of low water potential on Pg (Gao et al., 2002). Although these two approaches can simulate the effects of water stress, none of them consider the effect of Pe on gs, i.e. the mechanical advantages of epidermal cells over guard cells (Franks et al., 1998). This implies that: (1.6)   egs mPPg    , Introduction 28 where  > 0 is a proportionality constant accounting stomatal size and density and m ≥ 0 (dimensionless) is the epidermal mechanical advantage. Buckley et al. (2003) dealt with this issue in a model (‘BMF’ model) based on leaf, plant and stomatal water relations. In addition, they proposed an alternative ‘hydroactive feedback’ hypothesis where guard cell osmotic pressure is actively regulated in proportion to Pe (which acts as a sensor for changes in leaf water status) and to the ATP concentration of photosynthesising cells (which acts as a sensor for light and CO2). Thereby, gs is linked to photosynthetic activity: (1.7) eeg P      , where  is a scaling factor and  is ATP concentration in photosynthesising cells that can be simulated using the model of Farquhar & Wong (1984). When combined with a steady state model for liquid phase water flow from the soil to the leaf (E = K (  s –  leaf), where E is transpiration rate, K is leafspecific hydraulic conductance, and  s and  leaf are soil and leaf water potentials, respectively), Equations (1.6) and (1.7) lead to: (1.8)   VPD        KK gs s, where  is bulk leaf osmotic pressure and  =  – m +1. Important features of this model have to be highlighted. The use of m enables a better simulation of gs variations with VPD, namely the transient opening with increasing VPD which results from rapid, hydropassive responses, and the subsequent closure associated with the slower, hydroactive, energydependent osmotic response. Moreover, abscisic acid (ABA) effects can be introduced in the model. Buckley et al. (2003) showed that  could be interpreted as the ratio of the specific rates of active ion uptake and passive ion efflux in guard cells. Considering that the flux of ions entering the guard cells is ABA dependent, one would expect   and therefore the parameter Chapter 1 29   to decline as ABA concentration increases. Moreover, the most interesting advantage is that the parameters used in this model have explicit physiological meaning. Reviews on this topic (Damour et al., 2010; Egea et al., 2011b) recognized the potential of this model, but noted the difficulty in applying it, due to its high number of parameters. More and more mechanistic models are expected to appear as new information is acquired, aiming to bridge the gap between mechanistic and empirical models. In the present Thesis (Chapters 2, 3 and 4), this model will be parsed in more detail and will be slightly simplified with the aim of validating it in the field under water stress conditions. It will be suggested as well as a powerful tool not only for prediction, but also as a generator of new working hypothesis. Resumen En la presente Tesis se han utilizado una amplia gama de técnicas experimentales y análisis de modelización para estudiar los mecanismos fisiológicos involucrados en el control de la transpiración y aplicar los conocimientos adquiridos en la optimización del riego de cultivos de árboles frutales. Dos enfoques principales fueron utilizados para enlazar el conocimiento fisiológico emergente con la búsqueda de un manejo adecuado del riego en estos cultivos: sensores basados en medidas en plantas y modelos con base fisiológica o mecanísticos. En el Capítulo 2 se evaluó la regulación de los mecanismos que subyacen al control de la transpiración en una plantación de olivos en seto bajo condiciones de riego deficitario, combinando (i) el modelo hidráulico basado en las conductancias hidráulicas tanto del suelo como de la planta (Sperry et al., 1998), el modelo hidromecánico del control del estoma por balance hídrico (Buckley et al., 2003, o modelo ‘BMF’) y (iii) sondas de flujo de savia para validar los modelos con medidas independientes. En el Capítulo 3 se llevó a Chapter 2 Towards a more mechanistic model of water use: A case study in olive This Chapter is based on the published manuscript: Diaz-Espejo A, Buckley TN, Sperry JS, Cuevas MV, de Cires A, Elsayed-Farag S, MartinPalomo MJ, Muriel JL, Perez-Martin A, Rodriguez-Dominguez CM, Rubio-Casal AE, Torres-Ruiz JM, Fernández JE. 2012. Steps toward an improvement in process-based models of water use by fruit trees: A case study in olive. Agricultural Water Management 114: 37–49. Chapter 2 39 Introduction The main challenge for precision agriculture in fruit trees in arid and semiarid environments is the optimal management of irrigation. Two main approaches are being widely used with that purpose: the crop coefficient approach, also known as the FAO-56 approach, based on the PenmanMonteith equation to calculate the atmospheric demand and on the use of crop coefficients adapted to the orchard conditions (Allen et al., 1998); and the use of plant-based methods for a precise monitoring of the trees’ water stress (Fernández & Cuevas, 2010). In the first case, large uncertainties arise when equations are applied in different scenarios where soil conditions, phenological stages, orchard age, etc. vary (Pardossi and Incrocci, 2011). On top of that, the use of monthly values of the crop coefficient limits the temporal precision of this approach (Fernández & Moreno, 1999). The need for deficit irrigation in most fruit trees orchards (Fereres & Soriano, 2007; Ruiz-Sanchez et al., 2010) has driven substantial development of plant-based methods in recent decades. These methods can be used for the continuous and automatic monitoring of plant water stress, so they have a potential for high precision irrigation. A variety of sensors are used, including those related to measurements of sap flow (Fernández et al., 2008b), trunk diameter variations (Fernández & Cuevas, 2010; Ortuño et al., 2010), leaf turgor pressure (Fernández et al., 2011b), water content in the trunk (Nadler & Tyree, 2008), electric water potential (Oyarce & Gurovich, 2011) and canopy temperature (Jones, 1999). Plant-based methods have the potential advantage of measuring the plant’s response to the prevailing environmental conditions. The outputs, therefore, have a physiological basis, although their interpretation and application present challenges. Using these outputs to generate a measure of the degree of stress suffered by the plant, in order to apply management decisions, remains the main challenge. However, another approach has received some attention as well: physiologically based models. Modeling plant transpiration requires Process-based models of water use in olive 40 successful modeling of stomatal conductance, especially in fruit trees with a low decoupling factor (Jarvis & McNaughton, 1986) where transpiration is effectively controlled by stomata. Models of stomatal conductance have been approached from empirical (Jarvis, 1976; Ball et al., 1987; Leuning, 1995) and mechanistic points of view (Jarvis & Davies, 1998; Dewar, 2002; Gao et al., 2002; Buckley et al., 2003; Peak & Mott, 2011). However, the main challenge remains incorporating the effect of water stress in the models (Damour et al., 2010; Egea et al., 2011b). In addition to the importance of stomatal conductance, another key variable usually ignored in modeling fruit tree transpiration, and directly related to water stress and drought, is soil and plant hydraulic conductivity. Soil and xylem conductivity both decrease under hydraulic tension, and these decreases can be described well with physically based ‘unsaturated conductivity curves’ (for soil) or ‘vulnerability curves’ (for xylem). Improving the representation of soil and xylem hydraulics in models of crop water use is necessary to achieve a mechanistic link between soil water availability and canopy water use (Sperry et al., 2002). The present study will assess the ability of two process-based models – the Sperry et al. (1998, hereafter SACC) model of hydraulic limits and the Buckley et al. (2003, hereafter BMF) model of stomatal conductance – to inform management of water use in a hedgerow olive orchard. Our results demonstrate the central role played by the rhizosphere in the hydraulic limitation of transpiration in this orchard. This limitation can be partially managed by farmers if leaf area (pruning practices) and number of drippers in the irrigation system are modified. Moreover, our results show how much both hydraulic and non-hydraulic signals are involved in the control of transpiration by stomata. Chapter 2 41 Material and methods Study site and orchard characteristic The experiments were made in 2011 at the Sanabria orchard, a hedgerow olive commercial orchard at 25 km to the west of Seville, southwest Spain (37°15´N, −5°48´W). The trees, 5-year-old Olea europaea L. ‘Arbequina’, were planted at 4 m × 1.5 m (1667 trees ha−1). They had a single trunk with branches from 0.6 to 0.7 m above ground and the rows, 2.40 m tall and 1.96 m wide, run N–NE to S–SW. The area has a Mediterranean climate, with a mild, wet season from October to April and a hot, dry season for the rest of the year. Yearly average precipitation (P) and potential evapotranspiration (ETo) are 525.9 mm and 1542.4 mm, respectively (period 2002–2011). The experimental design was a randomized complete block design with four 12 m × 16 m plots per treatment. Each plot contained 8 central trees surrounded by 24 border trees. All measurements were made on the central trees of each plot. Two irrigation treatments were imposed in the orchard: a Control treatment where irrigation fulfilled tree water demand; and a regulated deficit irrigation treatment in which only 30% of the water added to control was applied (30RDI). The irrigation amounts (IA) supplied to this treatment varied according to the sensitivity to water stress of the crop at each phenological stage. Daily irrigation to replace 100% of the irrigation needs (IN) was calculated as IN = ETc − Pe, where ETc is the crop evapotranspiration and Pe the effective precipitation (75% of P recorded by the weather station in the orchard). Daily ETc values were calculated as ETc = Kc Kr ETo, where Kc is the crop coefficient and Kr is a coefficient related to the percentage of ground covered by the crop. We used the Kc values derived by Fernández et al. (2006a) for an orchard of similar characteristics, with a slightly greater canopy volume than that of the present orchard (0.76 in May; 0.70 in June; 0.63 in July and August; 0.72 in September; 0.77 in October; 1.07 in November) and we calculated a Kr value of 0.75 after Fereres & Castel (1981). Daily values of the FAO-56 Penman-Monteith ETo Process-based models of water use in olive 42 were collected from a nearby standard weather station belonging to the Agroclimatic Information Network of the Junta de Andalucía. Water for irrigation was supplied by a system consisting of one drip line per tree row with a 2 L h−1 dripper every 0.5 m with, and one caudalimeter per treatment to record the applied IAs. We used an irrigation controller (Agronic 2000, Sistemes Electrònics PROGRÉS, S.A., Lleida, Spain) for supplying the calculated INs. From June 7 to June 13, all trees in the orchard received enough water to match the crop water requirements. From June 14, day of year (DOY) 165, to October 24 (DOY 297), the 30RDI treatment was imposed in the orchard. All trees were fertilized by injecting a 8N-3P-8K + 0.05 % B + 0.05 % Fe solution into the irrigation system, once a week throughout the irrigation season. The amounts of fertilizers were enough to cover the tree requirements. Root and leaf area measurements To study root distributions, two soil trenches 1.5 m wide, 1.5 m deep and 3.5 m long were opened in July 2011, one in the Control treatment and one in the 30RDI treatment. Trenches were dug with a backhoe, and then the faces were squared and smoothed with a shovel. Using a 1 m × 1 m grid fixed on the trench face with nails, intersection root density (number of intersections per unit area) was counted in each grid unit of the trench wall face. A water spray bottle was used to highlight the roots to facilitate the identification of root ends. After counting all roots, a total of 17 soil samples with size 200 cm3 were removed by soil coring. Samples were transferred to the laboratory, where roots were washed from the soil samples and analyzed. All the roots were scanned with a WinRhizo LA 1600 scanner with a resolution of 300 dots per inch, and analyzed with the WinRhizo software (Régent Instruments, Quebec, Canada). Distribution of roots diameters and root length were measured. Root length density (RLD) was calculated from root length and soil volume sampled. Chapter 2 43 Leaf area (AL) was measured in each plot on the same days that the water status of the trees was monitored, i.e. once every two weeks during the irrigation season. Measurements were made at dawn with a LAI-2200 Plant Canopy Analyzer (Li-Cor, Inc., Lincoln, NE, USA). We followed the measurement strategy proposed by Villalobos et al. (1995) for olive orchards. Briefly, eight points per plot were measured in each of the four plots per treatment. In each plot, four points were measured just underneath the tree row, where LAI is maximum (LAImax), and other four points were measured in the midpoint between two rows where LAI is minimum (LAImin). The average LAI (LAIavg) was calculated using the fraction of ground cover (GC) as a weighting factor (LAIavg= LAImax GC+ LAImin (1-GC)). The average tree AL was calculated as LAIavg multiplied by the ground area per plot and dividing by the number of trees in the plot. Soil water status and physical properties In every plot we installed two access tubes for a Profile probe (Delta-T Devices Ltd, Cambridge, UK) at 0.5 m from the tree trunk and 0.1 m and 0.4 m, respectively, from the nearest dripper. Measurements of volumetric soil water content (θv, m3 m−3) in each access tube were made 1 - 2 times per week, at 0.1, 0.2, 0.3, 0.4, 0.6 and 1.0 m depths. The Profile probe was calibrated in situ, by comparing the values derived from the Profile probe readings with θv values measured with TDR probes (TDR FOM/mts, Institut of Agrophysics, Lubin, Poland). Cores of undisturbed soil were extracted from 0.1 to 1.0 m depth in cylinders (2.5 cm long and 5 cm diameter) to determine textural characteristics and dry bulk soil density (  ). Two contrasted soil layers were identified. This soil has a 0.6 m deep top layer with an average textural composition of 77.7 % sand, 2.2 % silt and 20.1 % clay, and  = 1.73 kg m−3. Below 0.6 m there is a less porous soil layer with average textural values of 60.9 % sand, 2.0 % silt and 37.1 % clay, and  = 1.82 kg m−3. Process-based models of water use in olive 44 Undisturbed soil cores were used to determine θv at −0.033 MPa using a 0.1-MPa porous ceramic pressure plate (Soil Moisture Equipment Corp., Santa Barbara, CA, USA) and −1.5 MPa by using a 1.5-MPa porous ceramic pressure plates with compressed air (Soil Moisture Equipment Corp., Santa Barbara, CA, USA). Values of residual θv (θresid), saturated θv (θsat), retention curve parameters (  , n) and soil hydraulic conductivity at saturation (Ksat) were estimated by using the Rosetta model (Schaap et al., 2001).  is related to the inverse of the air entry suction and n is a measure of the pore-size distribution. Rosetta model is based on van Genuchten model. Inputs to Rosetta were textural characteristics,  , and  v at −0.033 MPa and −1.5 MPa. Plant water status The time course of tree water status was monitored by measuring the leaf water potential at predawn (  pd) and midday (  md), and the midday stem water potential (  stem), once every two weeks during the whole irrigation season. Measurements were made with a Scholander-type pressure chamber (PMS Instrument Company, Albany, Oregon, USA) on one leaf per tree from two representative trees per plot (n = 8). For  pd and  md we sampled the 4th or 5th leaf below the apex of peripheral twigs at about 1.5 - 1.9 m above ground. They were healthy, fully developed, sun-exposed leaves facing east. For  stem we sampled leaves from the inner part of the canopy. These leaves were wrapped in aluminum foil ca. 2 h before the measurements to ensure hydraulic equilibration with stem xylem water. Osmotic pressure (  ) was measured on the same days as water potentials. At dawn, five leaves per plot were sampled, cleaned with a damp paper towel, packed in aluminum foil and immediately frozen in liquid nitrogen. These leaves were stored in a freezer until analysis. The expressed sap from each leaf was extracted according to Callister et al. (2006).  was then determined with a thermocouple psychrometer with six standard C-52 Chapter 2 45 sample chambers (Wescor Inc., Logan, UT, USA) connected to a datalogger (PSYPRO, Wescor Inc.). The measurements were carried out under constant temperature conditions. For each sample, one paper disc soaked with 10 l of the expressed sap was loaded into the sample chamber. A waiting time of 15 min was determined for sample equilibrium. Xylem vulnerability Xylem vulnerability was studied in current year olive shoots. The vulnerability to xylem cavitation was determined by the bench-top technique (Tyree & Dixon, 1986; Sperry & Tyree, 1988). Briefly, 1.5 m long branches were sampled under water from different representative olive trees, wrapped in plastic bags with wet paper towel inside to prevent water loss and transported to the lab where they were left to dry out on the bench. During the drying process, repeated measurements of xylem water potential and percentage loss of conductivity (PLC) were made. For PLC measurement, 30-mm long segments were sampled under water from the current-year shoots of the collected branches and connected to a XYL´EM apparatus (Bronkhorst, Montigny les Cormeilles, France) for determining their hydraulic conductivity (K). K was determined with a filtered (0.22 m) 50 mM KCl solution at 3 kPa until a steady-stated K was attained. Segments were then flushed at 150 kPa for 20 min to remove the embolisms and K measured again for determining the maximum K (Km). The PLC was then calculated as: (2.1)         m 1100PLC K K. Leaf gas exchange and sap flow Measurements of stomatal conductance (gs) were made with a Li-6400 open flow single pass gas exchange system using a standard 2 cm  3 cm leaf chamber (Li-Cor Inc., Lincoln, NE, USA) at the hours of maximum Process-based models of water use in olive 52 Fig. 2.1 G-I. Another evidence for the absence of roots below 0.45 m can be inferred from the constant value of  v, close to saturation, measured all along the season. This value did not decrease even when  v at 0.4 m presented values as low as 0.12 m3 m−3. An upper limit of 0.1 m was also set on the view of the low values measured at this shallow depth, close to residual. Values of  v were turned into  s by using soil physical parameters shown in Table 2.1. A single value of  s was obtained from the integration of  v from 0.2 to 0.4 m, and compared to the values of  pd measured in those trees, obtaining a good agreement (see Extra Fig. 2.1 in Appendix I). Fig. 2. 2 . Relationship between root length density (RLD) and root intersection on the trench wall per unit of surface. Data correspond to both Control and 30RDI treatments. r2 = 0.65 was significant at a value of P < 0.01. Despite significant differences of AL between treatments at the beginning of the growing season, the high growing rate of 30RDI trees Root counts cm-2 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 RLD (cm cm-3) 0.0 0.2 0.4 0.6 0.8 1.0 data slope= 2.47 Chapter 2 53 during spring allowed them to catch up Control trees by the time the daily irrigation was still on (Fig. 2.4). At this time of the year AL was on average 7.25 m2 per tree, meaning a LAI in the orchard of 1.43. However, Control trees grew again from mid July, up to maximum value of 12.3 m2 in October. This growing period was not observed in the 30RDI trees, which showed a constant AL all over the studied period. Assuming no changes in AR in 30RDI during this period, AR:AL was calculated as 0.38. Fig. 2. 3 . Spatial distribution of root length density (RLD ) on a trench wall in both the Control and 30RDI treatment. RLD values were calculated from the relationship shown in Fig. 2.2. Vertical bars on the upper horizontal line represent the tree trunks. RLD (cm cm-3) Distance (m) 0.5 1.0 1.5 2.0 2.5 3.0 Depth (m) 0.1 0.2 0.3 0.4 0.5 0.6 0 0.25 0.5 0.75 1.0 1.25 1.5 1.75 2.0 Control RLD (cm cm-3) Distance (m) 0.5 1.0 1.5 2.0 2.5 3.0 Depth (m) 0.1 0.2 0.3 0.4 0.5 0.6 0 0.25 0.5 0.75 1.0 1.25 1.5 1.75 2.0 30RDI Process-based models of water use in olive 54 Fig. 2. 4 . Seasonal courses of tree leaf area ( A L ) for the two irrigation treatments. DOY = day of year. Plant water status, leaf gas exchange and transpiration Full water availability in the soil for the Control trees was indirectly assessed on the view of the high values of  pd measured along the growing season (Table 2.2). However, 30RDI trees presented a minimum value of −1.5 MPa on DOY 209. Proportionally,  md showed a clear difference between treatments as soon as the irrigation frequency changed to once or twice per week in 30RDI, getting to a minimum value of −2.96 MPa on July 28, DOY 209 (Table 2.2). At the same time, gs showed nearly constant maximum values in Control trees along the season, meanwhile more than 3-fold lower gs values were measured in 30RDI trees at the end of the studied period DOY (60 = March 1) 60 90 120 150 180 210 240 270 300 AL (m2 tree-1) 0 2 4 6 8 10 12 14 Control 30RDI Chapter 2 55 (Table 2.2). These data matched well with Ep estimated from sap flow, shown in Fig. 2.5. This figure shows a period in which irrigation was applied daily to both treatments, although with slightly lower IAs to 30RDI, and the period in which the maximum stress was observed. During the first period, very close Ep were measured in both treatments, the small differences reflecting likely the different IAs mentioned above. However, a clear drop of more than 5-fold in 30RDI was observed during the second period plotted. In this period the Ep cycles of stress and recovery following irrigation were clear. Rp increased in both treatments during summer (Table 2.2). However, in Control trees Rp increased 1.4-fold, while in 30RDI trees the increase was over 8-fold. Table 2.2. Seasonal evolution of main physiological variables in both Control and regulated deficit irrigation trees (30RDI).  pd = pre-dawn leaf water potential;  md = midday leaf water potential; gs = stomatal conductance (mol m−2 s−1); Rp= plant hydraulic resistance (MPa mmol−1 m2 s). Each value represents the average of 8 replicates ± SE. Asterisk indicates significant differences between treatments on each date (t-student, P < 0.05). DOY Variable Treatment 165 (Jun 14) 181 (Jun 30) 209 (Jul 28) 223 (Aug 11)  pd Control −0.20 (0.04) −0.31 (0.05)* −0.37 (0.02)* −0.30 (0.03)* 30RDI − 0.15 (0.03) − 0.50 (0.03)* − 1.53 (0.15)* − 1.09 (0.11)*  md Control −0.99 (0.14) −1.33 (0.13) −1.37 (0.19)* −1.07 (0.12)* 30RDI − 0.77 (0.14) − 1.80 (0.23) − 2.96 (0.10)* − 2.50 (0.19)* g s Control 0.222 (0.01 ) 0.193 (0.01 )* 0.215 (0.02 )* 0.229 (0.02 )* 30RDI 0.233 (0.01 ) 0.098 (0.01 )* 0.069 (0.01 )* 0.069 (0.01 )* R p Control 0.53 (0.12) 0.59 (0.08)* 0.64 (0.12)* 0.74 (0.12)* 30RDI 0.47 (0.10) 2.03 (0.37)* 4.49 (0.82)* 4.10 (0.57)* Process-based models of water use in olive 56 Fig. 2. 5 . Time courses of air vapour pressure deficit (VPD) and transpiration estimated from sap flow measurements (Ep). Graphs on the left shows a period previous to the 30RDI treatment, when all trees in the orchard were daily irrigated. Graphs on the right correspond to days on which the 30RDI trees were irrigated twice per week. Arrows indicate irrigation events during this second period. DOY = day of year. SACC model Soil and plant hydraulic properties are the main determinants of the hydraulic limits for a species in a particular soil. Fig. 2.6A shows the relationship between Ks and  s based on the Van Genuchten equation. Despite of its high sand content, the orchard soil behaves more similar to a typical clay or silt soil, rather than a sandy soil. However, its high sand content makes it have 25-fold greater Ks than clay at  s = 0 MPa, and 13fold smaller Ks at  s = −1.0 MPa. Xylem vulnerability curves (Fig. 2.6B) VPD (kPa) 0 1 2 3 4 5 DOY (162= June 11; 208= July 27) 162 163 164 165 166 167 168 Ep (mmol m-2 s-1) 0.0 0.5 1.0 1.5 2.0 Ep Control Ep 30RDI 208 209 210 211 212 213 214 215 Chapter 2 57 show that olive is quite resistant to embolism. A PLC of 24 % was found at  stem = −3.0 MPa, the minimum  stem measured in the field site that year. The  stem at which 50 % of K is lost (P50) was −5.0 MPa. A complete loss of K (P100) was only achieved at values lower than −15 MPa. By comparison, grapevine  another woody crop considered to be well-adapted to semi-arid conditions  showed a P50 of −2.8 MPa, and a P100 around −5 MPa. Fig. 2.6. (A) Relationship between soil matric potential (  s) and soil hydraulic conductivity (Ks) for the Sanabria orchard soil at the top 0.6 m. The inset allows for a closer comparison of Ks at the highest  s. The small percentage of clay allows the maintenance of higher Ks than pure sand as  s decreases. Meanwhile the large percentage of sand of this soil allows for a high Ks at  s = 0. (B) Comparison of curves of vulnerability to cavitation of two species with contrasting resistance: Olea and Vitis. Data for Vitis was obtained from Choat et al. (2010). PLC data were fit with the Weibull equation: 1 −  ( ). Olea, d = 6.13, c = 1.81; Vitis, d = 2.97, c = 2.2. This information, together with AR and AL, was used to build the ‘envelope’ of water use proposed by Sperry et al. (1998, 2002). The difference between the actual transpiration and the envelope is called the safety margin, and it is reduced as  s is lower. Fig. 2.7 shows four different envelopes: two for olive and two for grapevine. In addition, actual data of B  stem (MPa) -14 -12 -10 -8 -6 -4 -2 0 PLC (%) 0 20 40 60 80 100 Olea Vitis A  s (MPa) -1.0 -0.8 -0.6 -0.4 -0.2 Ks (cm day-1) 100x10-12 1x10-9 10x10-9 100x10-9 1x10-6 10x10-6 100x10-6 1x10-3 10x10-3 100x10-3 1x100 10x100 100x100 1x103 Utrera Sand Silt Clay  s (MPa) -0.30 -0.25 -0.20 -0.15 -0.10 -0.05 0.00 Ks (cm day-1) 1e-6 1e-5 1e-4 1e-3 1e-2 1e-1 1e+0 1e+1 Process-based models of water use in olive 58 Ep vs  s were obtained from sap flow and continuous measurements of  v in the orchard and included in the plot for comparison. Data fit well under the olive envelope (thick line) for most of the range of  s. For very low values of  s data were above the limits. This could be due to uncertainty in  s arising from integration over three layers of heterogeneous soil moisture. A large safety margin is evident between the envelope at AR:AL measured in the orchard at high  s. However, this margin is severely reduced at  s close to −3 MPa. For grapevine, a similar value of AR:AL reduces the envelope notably, indicating that grapevine could not sustain the Ep values measured for olive. A three-fold increase in AR:AL increased the safety margin, emphasizing the differences in within-plant hydraulic limitation in the two species. Fig. 2. 7 . Results of plant transpiration ( E p ) simulated by the SACC model for two species of contrasting vulnerability to caviation: Olea, resistant, and Vitis, sensitive. All simulations were made for the soil conditions of our experimental orchard. Lines represent the hydraulic limit to transpiration fitted both by the rhizosphere and xylem. Simulation were made for two different root-leaf area ratios (AR:AL). The lower AR:AL value, 0.38, is the Bulk soil matric potential (MPa) -5 -4 -3 -2 -1 0 Ep (mmol m-2 s-1) 0.0 0.5 1.0 1.5 2.0 2.5 actual data Olea Olea, AR:AL = 0.38 Vitis, AR:AL = 0.38 Olea, AR:AL = 1 Vitis, AR:AL = 1 Chapter 2 59 actual value measured in the experimental orchard. Data points represent the actual E p derived from sap flow measurements in the orchard. The model can be used to simulate the effect of pruning intensity (changing AL) or the impact of the number of drippers (changes in volume of wet soil and therefore AR) in Ep. As expected, the increase in LAI decreases Ep (Fig. 2.8A), assuming that AR keeps constant (through changes in RLD, since the volume of soil is limited). If LAI is doubled, maximum Ep is halved. An increase in the number of wet drippers from 3 to 5 would nearly compensate for an increase in LAI. Since each dripper generates a wet bulb in the soil of 0.3 m in diameter, and each tree is 1.5 m apart from neighbors, five drippers is the maximum number that a line can hold. Fig. 2. 8 . Modeling exercise with the SACC model under well - irrigated conditions. ( A) Effect of increasing leaf area index (LAI) on transpiration rate (Ep). Close circles represent the actual values measured in our experimental orchard. We assumed a constant root length density when LAI increased. Simulation was made for three drippers (total volume of wet soil = 0.12 m3). (B) Effect of increasing number of drippers on Ep, for two different LAIs. We assumed a greater root area when the number of drippers, and therefore the volume of wet soil, increased. It was assumed that root distribution followed the pattern shown in Fig. 2.3. B # drippers 1 2 3 4 5 Ep (mmol m-2s-1) 0.0 0.5 1.0 1.5 2.0 2.5 LAI 1.5 LAI 3 A LAI (m2 m-2) 0.5 1.0 1.5 2.0 2.5 3.0 Ep (mmol m-2s-1) 0.0 0.5 1.0 1.5 2.0 2.5 actual value Process-based models of water use in olive 60 BMF model gc values estimated from sap flow measurements are plotted in Fig. 2.9 for the days leaf gas exchange and leaf water potential were measured (Table 2.2). The use of gc values allowed us to have complete series of diurnal evolution of a close surrogate of gs to apply the model. On DOY 165 (June 14) during the period of daily irrigation in 30RDI, no significant differences were observed between treatments. However, differences emerged as the soil dried out. A 4-fold decrease in 30RDI gc compared to Control gc values was observed on DOY 209 (July 28). The model fit well to measure gc on all dates shown and in both treatments. The model was able to reproduce the two peaks measured on at the beginning of DOY 181 (June 30). The seasonal evolution of the three parameters of the model is plotted in Fig. 2.10.  was measured and input to the model. Fig. 2.10A shows identical osmotic adjustment (change in  with  s) in both treatments despite their differences in  md. Tree hydraulics showed differences between treatments (Fig. 2.10B). These differences were evident from DOY 181, and especially later on, when a 3-fold increase in R of 30RDI was estimated by the model (which was fitted by least squares). Control trees showed a nearly steady value of R along the season.  (also fitted by least squares) showed a similar seasonal trend in both treatments (Fig. 2.10C), except on DOY 223 (August 11) when a small recovery was obtained in Control. The model always predicted higher values for Control than for 30RDI. Chapter 2 61 Fig. 2. 9 . Simulation of the evolution of canopy conductance ( g c ) by the BMF model, for both Control (close circles) and 30RDI (open circles) trees. Points represent actual gc data derived from sap flow measurements made in the orchard every half-hour. The lines represent the simulated values. The thicker line on DOY 165 shows the result of the model with a variable soil matric potential  s) as mentioned in the text. Similar results were obtained for the 30RDI trees on that day, but the line has not been included for clarity. The input meteorological data was obtained from the weather station located in the orchard, and  s was estimated from predawn leaf water potential. August 11 (DOY 223) 05:00 09:00 13:00 17:00 July 28 (DOY 209) GMT 05:00 09:00 13:00 17:00 gc (mol m-2 s-1) 0.00 0.02 0.04 0.06 0.08 0.10 Control data Control modelled 30RDI data 30RDI modelled June 14 (DOY 165) gc (mol m-2 s-1) 0.00 0.02 0.04 0.06 0.08 0.10 0.12 June 30 (DOY 181) Process-based models of water use in olive 68 the value of 0.38 estimated. We think, however, that the bias is not important since the volume of soil for active growing is limited to wet bulbs, where we have measured RLD values close to the maximum values reported by olive orchards under drip irrigation (Fernández et al., 1991). The effect of greater xylem vulnerability can be evaluated with grapevine simulations (Figs. 6 and 7). Grapevines planted at Sanabria and with an AR:AL = 0.38 would not be able to maintain the same Ep as olive. Assuming a potential Ep of 1.3 mmol m−2 s−1, similar to that for olive in this study, and considering that the safety margin is maintained at high  s despite the decreasing water use envelope (Sperry et al., 2002, Fig. 2.3), maximum Ep would be reduced by half. Increasing AR:AL to 1.0 eliminates this limitation at high  s, and allows grapevine to approach the extraction limit of around −5 MPa imposed by the xylem. Changes in AR:AL could be achieved not only by increasing AR but also by decreasing AL. Some species like Quercus canariensis and Q. faginea, typical of our latitudes, are semi-deciduous, i.e. their leaves drop in summer, which reduces AL. In some species adapted to arid environments the reduction of AL in summer can be extremely severe, provoking leafless branches (Miranda et al., 2010). Contrary to what happens in plants under natural conditions, farmers can modify the AR:AL ratio by changing both AR and AL. Fig. 2.8 shows a simulation of the impact of a changing LAI on Ep if AR is kept constant. Ep is decreased by 50 %. In olive, as in most crops, yield is directly related to water transpired by the plant (Moriana et al., 2003; Fereres & Soriano, 2007). The simulation rests on two assumptions. One is that AR is maintained, which is likely due to the limitation of volume of wet soil imposed by localized irrigation and the high RLD already found. The second assumption is that the increment in Ep with increasing LAI is not dependent on a different percentage of sunny and shaded leaves. We did not consider, in the simulation, that increasing LAI enhances the number of shaded leaves vs sunny leaves. But if that effect is taken into account, the Chapter 2 69 assumption also holds, as Fernández et al. (2008a) demonstrated using a multilayer radiation interception model. The simulation in Fig. 2.8 shows an important impact of the pruning management in the orchard, and sets an optimal LAI value for an irrigation system with three drippers per tree. However, larger yield will be obtained from larger LAI, since more shoots are able to carry fruits. On the other side, the reduction in Ep due to the increasing LAI could have a negative impact on yield. Currently, we have no answer to this trade-off. However, farmers can influence AR to compensate for increasing values of AL. The question is: how many drippers do we need to install per tree to recover Ep (to that at previous values of LAI) if we double LAI? The answer can be assessed by the model changing the volume of soil wetted by a dripper and considering steady RLD. The double LAI is nearly compensated by increasing the number of drippers from three to five. A model for the actual transpiration We have seen how the SACC model can predict hydraulic limits based on plant and soil characteristics, and how it can be applied to orchard management. However, this model was not designed to predict diurnal courses of actual transpiration, unless the diurnal course of leaf water potential is input as well. Additional insights can be gained by simulating stomatal responses to atmospheric demand and soil water deficit. In canopies well coupled to the atmosphere most of the transpiration is driven by VPD (Jarvis & McNaughton, 1986). This is the case of olive tree orchards (Moreno et al., 1996; Tognetti et al., 2009). An important characteristic of well coupled canopies is that gs exerts a strong control of transpiration. This means that if we can model gs satisfactorily, we can predict Ep. Traditionally, the most widely used gs models have been those of Jarvis (1976) and Leuning (1995). However, these models, although able to mimic the stomatal response in most simulations, have difficulty simulating the effect of water stress (Vico & Porporato, 2008; Egea et al., 2011b). Process-based models of water use in olive 70 Several attempts have been made to consider the response of stomata to a drying soil, for instance by including the effect of ABA (Gutschick & Simonneau, 2002), which has been reported to act as a chemical signal from roots to leaves. Egea et al. (2011) found a suitable solution by including a soil moisture dependent function to account for the effects of water stress on gs. The hydromechanical model proposed by Buckley et al. (2003) (BMF) has not been widely applied yet and it has an advantage over most others in that its parameters have explicit physiological meaning. Recent reviews on this topic (Damour et al., 2010; Egea et al., 2011b) recognized the potential of this model, but noted the difficulty in applying it, due to its high number of parameters. Our results show we were able to apply and validate a slightly simplified form of this model. We estimated canopy conductance (gc) from sap flow data as a surrogate of gs (Fig. 2.9) in order to obtain long time series of this variable. Values of gc compared well with gs measurements made at the leaf level. In these same trees, gs measured in sunny and shaded leaves were on average 0.23 mol m−2 s−1 and 0.07 mol m−2 s−1, respectively. Assuming a fraction of sunny leaf area between 0.2 and 0.3 (based on measurements by Moreno et al., 1996, lowered from their value of 0.3 to account for mutual shading in hedgerows), the gc value obtained is very similar to the value estimated from leaf gas exchange. The BMF model fitted our data quite well in both treatments as soil dried out (Fig. 2.9). The model’s real strength, however, is that it allowed us to analyze the physiological parameters obtained (Fig. 2.10). In our case, only two parameters of three were fitted. The seasonal evolution of  shows that similar osmotic adjustment occurred in both treatments. This was especially strong in Control on DOY 209 since  md was only −1.37 MPa. Osmotic adjustment has been interpreted as a mechanism for maintaining leaf turgor pressure when  md is reduced to withdraw water from drying soils (Dichio et al., 2006). In the case of 30RDI this explanation fits perfectly.  md on DOY 209 was as low as −2.96 MPa Chapter 2 71 and  = −3.03 MPa. Therefore, another explanation is required for the osmotic accumulation observed in the Control treatment. One possibility is accumulation of photosynthates, which would build up a high  to facilitate the transport of assimilates to phloem. The theory of passive loading to phloem for the primary photoassimilates, sucrose and sugar alcohols, could explain this increase in  independently of a response to water stress (Rennie & Turgeon, 2009). Species accumulating mannitol, like olive (Flora & Madore, 1993), use this strategy preferentially, probably because it requires no energy (Reidel et al., 2009). This explanation agrees well with the activation of growing in Control trees from DOY 190 (July 9) (Fig. 2.4). Although this hypothesis deserves further experimental study, it might have important implications in the identification of the growth inhibition threshold for olive, and the determination of optimal water potentials for managing irrigation. One parameter fitted by the model is plant hydraulic resistance, R (Fig. 2.10B). The small increase in R observed in Control trees was expected, since  s did not exceed −0.4 MPa in the whole period, and  md was never lower than −1.4 MPa. On the contrary, 30RDI experienced a progressive increase in R, in agreement with the decrease in  s. The seasonal pattern of both treatments fits well to data of Rp shown in Table 2.2. Differences in absolute values of both variables also highlight the importance of the soil component included in R: these differences increased as soil dried out, as suggested by Sperry (2000). Similar values of Rp and its seasonal evolution have been reported previously for olive (Tognetti et al., 2009), as well as for other Mediterranean tree species like Quercus rotundifolia (David et al., 2004), but we have not found in literature such a marked difference between controls and stressed plants under field conditions as in this study. The likely explanation for this behavior is the small AR:AL in our olive trees. The reduced IAs and low irrigation frequency in 30RDI affected the hydraulic capacity of the plant. Despite this constraint in AR, Control plants did not Process-based models of water use in olive 72 experience important changes in their R. The daily irrigation with sufficient IAs was enough to maintain a nearly steady maximum Ep. However, it is important to note that the increase in R in 30RDI looks disproportionate to the PLC predicted by the vulnerability curve on Fig. 2.6. PLC on this figure refers only to shoot xylem vulnerability to cavitation, while R comprises the soil-to-leaf continuum. This means that the increase in R arose elsewhere in the plant, likely roots or leaves. Indeed, the vulnerability to cavitation has been reported to be higher in both roots (Alder et al., 1996; Kolb & Sperry, 1999) and leaves (Brodribb & Holbrook, 2003; Zufferey et al., 2011), than in the xylem. Vulnerability to cavitation in petiole and leaf lamina have been recognized in recent years to play an important role in stomatal regulation (e.g., Guyot et al., 2011). The other fitted parameter,  , represents the sensitivity of guard cells to changes in turgor pressure. Unlike R,  showed clear seasonal dynamics in both treatments (Fig. 2.10C).  includes the effect of guard cell solute efflux, so it should decline in response to hormonal signals from drying roots, like ABA (Buckley, 2005). This is consistent with the inferred seasonal patterns.  also includes the effect of stomatal size and density, which were similar between treatments in this study (data not shown). Our data suggest that the putative drought signal was stronger in 30RDI for most dates. On DOY 165  was one third as high in Control than in 30RDI, before declining to a minimum in both treatments on DOY 209 and then recovering on DOY 223. The main conclusion is that the stomatal sensitivity to leaf water status was regulated seasonally not only in 30RDI, but also in well-irrigated Control trees – suggesting that  is not primarily regulated by soil moisture (nor, by inference, by ABA signals from droughted roots) in this species. This is consistent with some research suggesting that ABA is synthesized primarily in leaves, rather than in roots (Holbrook et al., 2002; Christmann et al., 2005), and that the travel time of Chapter 2 73 ABA in woody species may be too great for it to serve as a rapid long distance signal (Perks et al., 2002). To sum up this section, application of the BMF model to our data suggests that stomata in olive are regulated seasonally by something other than purely hydraulic signals, and that this occurs not only in droughted but also in well-watered trees. These signals might be chemical (for example ABA, nitric oxide, reactive oxygen species, etc.; Jiang & Zhang, 2001; Neill, 2007), or physical (for example electric signals, Stahlberg et al., 2001; Oyarce & Gurovich, 2011). Additionally, hydraulic capacity of plants receiving only 30 % of water supplied to Control is severely affected under our conditions of climate, soil and irrigation strategy. In any case, the BMF model can serve as a useful research tool to understand the mechanisms behind observations, and as a platform to accommodate experimental knowledge from the literature. Conclusions The use of two process-based models helped us to advance understanding of water use by an olive orchard planted in hedgerow. The SACC model confirmed that the main limitation in the water use by olive trees in this orchard was in their rhizosphere. The limited volume of wet soil, determined by the number of drippers, reduced the ratio of root to leaf area. This reduction imposed a large hydraulic limitation to transpiration as bulk soil water potential decreased. The model was able to predict the impact of soil type, ratio of root and leaf areas on the limit of extraction of water by the plant. This has important practical implications for pruning and irrigation management, as the model can be used to assess the impact of changes in leaf area and number of drippers. Increasing the number of drippers from the actual three to five would be necessary to compensate for Process-based models of water use in olive 74 two-fold increment in leaf area if the goal were to keep maximum transpiration values. The BMF model simulated satisfactorily the actual canopy conductance on several dates through the summer, both in wellwatered and water stressed plants. Plants of both water treatments made similar osmotic adjustment. However, soil-to-leaf hydraulic resistance in stressed plants increased more than 4-fold during the summer. A potential involvement of regulating signals, other than purely hydraulics, was evident in both treatments, although our data suggests that these signals were themselves regulated by something other than soil water status. Chapter 3 The contribution of hydraulic and non-hydraulic factors to suppression of stomatal conductance during soil drought: A case study in almond. This Chapter is based on the manuscript under preparation and submitted to New Phytologist in September 2013: Rodriguez-Dominguez CM, Buckley TN, Egea G, de Cires A, Hernandez-Santana V, Diaz-Espejo A. The contribution of hydraulic and non-hydraulic factors to suppression of stomatal conductance during soil drought. Analysis of stomatal response to soil drought 84 darkness in 15 steps. Temperature and CO2 were 28-32 ºC and 390 mol mol−1, respectively. Maximum A (Amax, mol m−2 s−1), the curvature parameter (  , dimensionless) and maximum quantum yield of CO2 (  , the initial slope of A versus PPFD, dimensionless) were determined by least squares curve fitting to a non-rectangular hyperbola. Parameter values (  = 0.71 and  = 4  = 0.20 electrons photon−1, where  is the effective maximum quantum yield of electrons) were similar between treatments. ABA extraction, purification and quantification Predawn and mid-day leaves were sampled as for  measurements. Leaf ABA was measured by the liquid chromatography-electrospray/tandem mass spectrometry method of Gómez-Cadenas et al. (2002). Samples of ca. 400 mg of frozen leaf tissue, with midribs removed, were milled with liquid nitrogen and homogenized and extracted in 5 ml of distilled water. An aliquot of 50 L of 2-ppm deuterated abscisic acid (dABA) was added as an internal standard. Samples were centrifuged (26000 min−1; 7 min; 8 ºC), supernatants were acidified to pH 3.0 (150 L acetic acid 30 % (v/v)) and leaf extracts were 2-times partitioned with 3 mL of diethyl ether. Organic phases were collected in test tubes and totally evaporated providing a gaseous nitrogen flow. Tube walls were washed with 1 mL diethyl ether and desiccated again. Dry residues were re-suspended in 500 L methanol, completed to a total volume of 1 mL with Milli-Q quality (reverse osmosis) water and filtered through 25 mm diameter polypropylene membrane syringe filter (ø 0.2 m, VWR® International, Pennsylvania). Analyses were performed using an Agilent 1290 Infinity HPLC system (Agilent Technologies Inc., CA) coupled with an electrospray/tandem mass spectrometer (3200 QTRAP® LC/MS/MS System, AB SCIEX, Framingham, MA) and data were processed with mass spectrometry software (Analyst® Software, AB SCIEX). Leaf ABA was normalized by fresh weight (FW, g). Chapter 3 85 Statistical analysis We used linear mixed models to analyze effects of irrigation treatment (as a fixed factor) on SWC,  leaf, gs, K, n, π, leaf ABA concentrations, Vc,max, Jmax and gm. We used random factors when necessary to describe our experimental design (leaf within pot for  leaf and gs, and soil water content probe within pot for SWC). Variables were transformed to improve normality or to fix non-constant variance in residuals when needed. Models were fitted by restricted maximum likelihood (REML) in R (package 'nlme R’; Pinheiro et al., 2012). Multiple-comparison analyses were conducted when an overall significant effect was detected (  = 0.05). BMF model of stomatal conductance We used a modified form of the stomatal conductance model originally presented by Buckley et al. (2003) (hereafter, the BMF model) to examine the mechanistic basis of observed changes in gs. This model is based on leaf, plant and stomatal water relations, and the hypothesis that guard cell osmotic pressure is actively regulated in proportion to epidermal turgor pressure (which acts as a sensor for changes in leaf water status) and to the ATP concentration of photosynthesising cells (which acts as a sensor for light and CO2). In the Appendix II, we present the model in greater detail, and we derive the following modified form of the model: (3.1)   w na K ΨnaK g Δ s s    , where K is leaf-specific hydraulic conductance,  s is soil water potential,  is bulk leaf osmotic pressure and w is leaf to air water vapor mole fraction gradient. n and a capture non-hydraulic effects: a is mesophyll ATP concentration expressed relative to its maximum value and n is a lumped parameter representing other non-hydraulic factors: Analysis of stomatal response to soil drought 86 (3.2) m   n, and (3.3) m    a, where  is ATP concentration in photosynthesising cells. The parameters embedded in n are  , a proportionality factor that scales guard and epidermal cell turgor pressures to gs;  , a proportionality factor that scales the product of  and epidermal turgor to changes in guard cell osmotic pressure; and  m, the maximum  (the total pool of adenylates, ADP + ATP). We simulate  with the model of Farquhar & Wong (1984), which was derived from the photosynthesis model of Farquhar et al. (1980) and is presented in detail in the Appendix II. Buckley et al. (2003) showed that  could be interpreted as the ratio of the specific rates of active ion uptake and passive ion efflux in guard cells. Because abscisic acid (ABA) affects guard cells by stimulating passive efflux (Hetherington, 2001), one would expect  – and therefore the parameter n in the modified model – to decline as ABA concentration increases. Equation (3.1) was derived from the BMF model based on several assumptions, namely that the osmotic gradient from guard to epidermal cells, not to the apoplast, is the target for active regulation; that the resistance from epidermal to guard cells is negligible compared to the resistance from the soil to the epidermis; that epidermal and bulk leaf osmotic pressure are similar; and that the response of gs to PPFD is homogeneous (Extra Fig. 3.1 in Appendix II). These assumptions are discussed in greater detail in the Appendix II. Chapter 3 87 Attributing changes in stomatal conductance to factors in the BMF model We parsed changes observed in gs between 21 and 31 August in the WS treatment into changes due to individual factors in Equation (3.1) using the following expression: (3.4) w w ggg K K g a a g n n g gsΔ ss s ssss s                         . The percent contribution of a factor was computed by dividing it by the sum of all of contributions and multiplying by 100 (details in Appendix II). Stomatal limitations Jones (1985) and Grassi & Magnani (2005) proposed delineating the limitations on photosynthesis in terms relatives partial derivatives of A with respect to various limiting factors, which has proven useful for attributing differences in photosynthesis to those factors. The same concept can be applied to Equation (3.1) to compute the hydraulic and non-hydraulic limitations to stomatal conductance,  h and  nh, respectively. We define the hydraulic limitation to be that due to investments (or lack thereof) in hydraulic conductance, K (  h = lngs/lnK), and the non-hydraulic limitation as that due to limitations in the product na (  nh = lngs/ln{na}). These derivatives are: (3.5) nawK na K g     Δln ln s h  , and (3.6) h s nh 1 Δ Δ ln ln       nawK wK na g. Note that if one chose instead to define  nh in relation to either n or a alone, the same expression would arise: Analysis of stomatal response to soil drought 88 (3.7) nawK wK a g n g        Δ Δ ln ln ln ln ss . Results Soil water content (SWC) was similar between the well-watered (WW) and water-stressed (WS) treatments during the first two measurement cycles (on 15 and 21 Aug), but declined significantly in WS during the third cycle (31 Aug) as compared to WW (Fig. 3.1A - C). Evaporative demand (w) was greatest on 21 Aug (exceeding 70 mmol mol−1), but was also quite high during the third cycle (Fig. 3.1G - I). Despite the high w on 21 Aug, leaves in both treatments were able to maintain stomatal conductance at levels similar to 15 Aug, when w was much lower (Fig. 3.2B). However, the lower SWC in WS on 31 Aug led to a large decline in both gs and predawn leaf water potential (Fig. 3.2A). Chapter 3 89 Fig. 3. 1. Time courses of soil water content (SWC) measured on the experi mental dates, and photosynthetic active photon flux density (PPFD), air temperature (Ta) and leaf to air water vapor mole fraction gradient (w) on three experimental periods. Central dates of each panel correspond to daily measurement dates (August 15, 21 and 31). SWC is represented for the two irrigation treatments supplied: well watered treatment (WW) and water stressed treatment (WS). Error bars on SWC represent standard errors, n = 6. Asterisks indicate statistically significant differences between treatments (P < 0.001). 00:00 12:00 00:00 12:00 00:00 12:00 Ta (ºC) 15 20 25 30 35 40 PPFD (mol m-2 s-1) 0 200 400 600 800 1000 1200 1400 1600 SWC (m3 m-3) 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 WW WS 00:00 12:00 00:00 12:00 00:00 12:00 00:00 12:00 00:00 12:00 00:00 12:00 00:00 w (mmol mol-1) 0 10 20 30 40 50 60 70 80 Aug 14 Aug 15 Aug 16 Aug 20 Aug 21 Aug 22 Aug 30 Aug 31 Sept 01 ABC DE F GHI **** Date and Time (h, GMT) Analysis of stomatal response to soil drought 90 Fig. 3.2. Diurnal courses of (A) leaf water potential (  leaf) and (B) stomatal conductance (gs) on the three experimental dates. Error bars show standard errors for n = 6 (  leaf) and n = 8 (gs). Asterisks indicate statistically significant differences between treatments on August 31 (* = 0.01 < P < 0.05, ** = P < 0.01). GMT = Greenwich Mean Time. Therefore, we focused on what caused the decline in gs in the WS treatment between 21 and 31 Aug. We investigated the physiological causes of this decline by a combination of measurements and inference. The measurements included predawn leaf water potential as a proxy for soil water potential (  s), leaf osmotic pressure (  ), PPFD, w, and leaf abscisic acid concentration ([ABA]) at predawn and mid-day. The inferential Time (h, GMT) 05:00 09:00 13:00 17:00 gs (mol m-2 s-1) 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 B  leaf (MPa) -3.5 -3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 0.5 WW - Aug 15 WW - Aug 21 WW - Aug 31 WS - Aug 15 WS - Aug 21 WS - Aug 31 A * * ** ** ** ** ** ** ** ** ** Chapter 3 91 approach entailed fitting the BMF model (Equation 3.1) to our data to infer changes in two fitted parameters – leaf hydraulic conductance (K) and the lumped parameter of non-hydraulic factors (n) – and using the model to separate the contributions of each factor in the model to the observed decline in gs. We expected that n would decline but K would remain constant during this decline, so that n would emerge as the major driver of declining gs. Our rationale was twofold: first, diurnal minimum leaf water potential was similar between treatments and days (Fig. 3.2A), suggesting any additional loss of hydraulic conductivity on 31 Aug would be minimal; and second, the parameter n contains embedded within it the parameter  (Equation 3.2), which captures the effect of passive, outward-rectifying osmotic solute loss from guard cells – the main process believed to mediate stomatal closure in response to the biochemical drought signal ABA (see Appendix II). Analysis of stomatal response to soil drought 92 Fig. 3. 3. Evolution of g s data (points) and g s fitted with the BMF model (lines) on the three experimental dates during the study. For gs data we used n = 8. Error bars show standard errors. The model was able to reproduce diurnal variations in gs across the study period, as well as the decline in gs in WS on 31 Aug (Fig. 3.3). Trends in  s, PPFD and w differed:  s was lower in WS on 31 Aug than in other treatments and days (Fig. 3.2A), but neither PPFD, w nor  differed Time (h, GMT) 07:00 11:00 15:00 19:00 0.0 0.1 0.2 0.3 gs (mol m-2 s-1) 0.0 0.1 0.2 0.3 0.0 0.1 0.2 0.3 0.4 WW data WW modelled WS data WS modelled August 15 August 21 August 31 A B C Chapter 3 93 significantly between 21 and 31 Aug (Fig. 3.1B, C; Fig. 3.4C). K was similar between treatments on 21 Aug but declined on 31 Aug by 54 % in WS and 22 % in WW (Fig. 3.4B). Conversely, the non-hydraulic term n declined on 31 Aug, as expected, but by a greater degree in WW (71 %) than in WS (26 %), which contradicted our expectations. To assess whether these variations in n were paralleled by changes in biochemical drought signals, we compared them to trends in leaf [ABA]. Mid-day [ABA] was similar in all cases except WS on 31 Aug, when it increased (Fig. 3.5B). Conversely, predawn ABA was greater in WW than in WS on 21 Aug (Fig. 3.5A), despite the similar conditions between both treatments, but this pattern was reversed on 31 Aug, when pre-dawn ABA was greater in WS than in WW (Fig. 3.5A). Overall, gs was negatively correlated with [ABA] (Fig. 3.6A, B), as expected, but the parameter n was not (Fig. 3.6C, D), contrary to expectations. Analysis of stomatal response to soil drought 100 constrained to the ‘binary’ approach of competing alternative hypotheses, and that in fact a more integrative approach is more informative. Response to water stress in almond The imposition of soil water deficit on WS plants produced a large decrease in gs on 31 Aug which reduced transpiration enough to maintain minimum daily  leaf at a constant value of around –2.7 MPa, similar to the value observed before soil drought. We found that leaf [ABA] at mid-day increased coincident with the decline in gs, and that [ABA] at predawn was negatively correlated to  s. Similar results have led other authors to suggest that under soil water deficit, roots synthesise ABA, which is delivered in the transpiration stream to guard cells where it induces ion efflux, thus reducing stomatal aperture (Tardieu & Davies, 1992; Dodd et al., 2010). It has been argued that such signals are necessary because a negative feedback response of gs to  leaf cannot produce true homeostasis in  leaf. A counter-argument is that homeostasis could result if the gain of the negative feedback were amplified by a reduction in hydraulic conductance (Oren et al., 1999; Buckley & Mott, 2002b; Buckley, 2005). Although we did observe a 35% decline in hydraulic conductance coincident with the decline in gs, it is not self-evident whether this should be sufficient to produce isohydry. Therefore, we developed a novel and formal approach, based on the stomatal conductance model of Buckley et al. (2003) (the BMF model), to address this question. Despite the recognized potential of this model (Damour et al., 2010; Egea et al., 2011b), it has not been widely applied in field experiments, partly because it has many parameters that are difficult to estimate. To overcome this limitation, we simplified the model to produce a modified form with just three parameters (Equation 3.1), which fitted our data reasonably well. We then used a differential analysis of Equation (3.1) to estimate the relative contribution of changes in K and ABA (the effect of which is embedded in the parameter n in Equation 3.1), Chapter 3 101 as well as other biophysical factors in the model, to the observed decline in gs. That analysis concluded that the decline in K was responsible for nearly half (46 %) of the decline in gs, whereas the decline in the ABA-related parameter n was responsible for only 8 %. Most of the remaining 44 % was attributable to the direct effect of reduced  s (38 %), with small effects due to  , w and irradiance. Why, then, did the increase in [ABA] at mid-day apparently not exert greater control over stomatal conductance? To help answer that question, we computed the relative limitations to stomatal conductance due to hydraulic vs non-hydraulic factors, symbolised as  h and  nh, respectively (Equations 3.5 and 3.6). These limitations are directly analogous to the relative photosynthetic limitations proposed by Jones (1985) and developed further by Grassi & Magnani (2005). An unexpected insight of this limitations analysis was that non-hydraulic factors generally exert very little control over gs during the middle of the day. The reason is that  nh is a decreasing function of irradiance and w (Equation 3.6:  nh = K/(K + naw), where a is the model parameter that increases in relation to irradiance, and n is the parameter that should decrease as [ABA] increases). Furthermore, any decrease in K that may occur during soil drought will further reduce the non-hydraulic limitation of gs. In light of these insights, it is less surprising that our model analysis attributed little control to mid-day [ABA]. There are several reasons to suppose this conclusion may not apply broadly to most species. Firstly, both w and PPFD are quite high in our study sites in southern Spain (e.g., w often exceeds 70 mmol mol−1), and this increases the control of gs by hydraulic factors (  h, Equation 3.5). Secondly, the greatest driver of the decline in gs in this study was a decline in K, yet because species differ widely in the vulnerability of K to reduced water potential, soil drought will affect K differently across species. Thirdly, Analysis of stomatal response to soil drought 102 the BMF model does not represent a universal consensus about the mechanisms of stomatal regulation. In particular, the nature of the link between the negative feedback response to water status and the effect of ABA is poorly understood (Buckley & Mott, 2013). The model assumes guard cells actively modulate osmotic pressure in relation to water status – an hypothesis that is strongly supported by a great deal of circumstantial evidence (Buckley, 2005) but has never been tested directly. Thus, although we feel that this model is currently the best tool available for mechanistic analysis of stomatal regulation, failure of any of its assumptions could mitigate the conclusion of our analysis. What is the role of ABA? Our analysis complements a range of evidence that hydraulic responses may explain much of the regulation of stomatal conductance in relation to soil drought in some species (Fuchs & Livingston, 1996; Comstock & Mencuccini, 1998; Yao et al., 2001). Nevertheless, we did observe large increases in leaf [ABA] at mid-day coincident with the decline in gs, and some speculation is warranted concerning the role of this increase. The hypothesis that root-derived ABA regulates gs during soil drought has been extensively studied and has even impacted agriculture, through irrigation strategies based on split-root-zone drying (Dodd et al., 2008). However, this hypothesis has been challenged in tall trees (Perks et al., 2002) and in experiments in which shoots were grafted onto rootstock deficient in ABA synthesis (Holbrook et al., 2002; Christmann et al., 2005). The latter authors concluded that most ABA synthesis was confined to the vasculature and guard cells. This was later confirmed and extended by Gálvez-Valdivieso et al. (2009), who found that most ABA synthesis was localised in the vascular parenchyma, where it activates a signalling network in neighbouring bundle sheath cells linked to hydrogen peroxide accumulation under high light. More recently, Pantin et al. (2012) showed that ABA reduces the conductivity of water flow pathways distal to the xylem, perhaps by Chapter 3 103 modulating aquaporin activity. Together, these results suggest an hypothesis: namely, that the decline in K in our experiments, which we concluded was responsible for nearly half of the decline in gs, was in fact mediated at least in part by ABA – i.e, the hydraulic and non-hydraulic limitations may be more closely coupled than previously thought. Conclusions We found that stomatal closure leading to isohydric behaviour under soil drought in almond could be explained by a combination of both hydraulic and non-hydraulic factors, but that nearly half of the decline in gs was attributable to a decline in hydraulic conductance, and over a third was attributable to the direct effect of reduced soil water potential. Our novel model-based approach to parsing changes in gs into contributions of various biophysical factors has the potential to yield more insight than a traditional binary approach based on mutually exclusive alternative hypotheses. Chapter 4 Combining a process-based model of stomatal conductance with leaf turgor pressure related probe measurements to study the regulation of plant water status and stomatal conductance under drought This Chapter is based on the manuscript under preparation: Rodriguez-Dominguez CM, Buckley TN, de Cires A, Fernández JE, Perez-Martin A, Diaz-Espejo A. 2014. Combining a process-based model of stomatal conductance with leaf patch pressure probe measurements to study the regulation of plant water status and stomatal conductance under drought. Chapter 4 107 Introduction Precision irrigation in horticultural crops is highly demanded by farmers for an optimal water management worldwide. The use of plant-based sensors is in many cases the recommended option since plants are an integral component of the soil-plant-atmosphere continuum (Jones, 1999; Jones, 2004; Fernández et al., 2008b; Nadler & Tyree, 2008; Fernández & Cuevas, 2010; Ortuño et al., 2010; Oyarce & Gurovich, 2011). These sensors have to meet a number of criteria to better approach precise irrigation scheduling: being a reliable and sensitive water stress indicator, user-friendly, minimally invasive and suitable for automatically and continuously data collection and transmission throughout a whole irrigation season the most relevant (Jones, 2004; Fernández, 2014a). The recently developed leaf patch clamp pressure probe (LPCP probe) has proved to meet those criteria to precisely monitor water stress in several plant species and crops (Westhoff et al., 2009; Rüger et al., 2010a; Fernández et al., 2011b; Lee et al., 2012; Bramley et al., 2013). Its output targets on turgor pressure (Rüger et al., 2010a; Ehrenberger et al., 2012a, 2012b), one of the physiological variables recognized to be among the most sensitive to water stress (Jones, 2004, 2007). Additionally, the applicability of leaf patch pressure probes has been extended to more specific studies in the regulation of physiological processes by the leaf water status (Ache et al., 2010; Bauer et al., 2013), increasing its potential to disentangle mechanisms of response to water stress. Besides that, process-based models have been also suggested as a powerful tool to improve our understanding of plant physiological mechanisms involved in the response to water stress (Buckley & Mott, 2013). These models use physiological based parameters and have the strength to mechanistically simulate leaf and plant behaviors. In the present work, we used a process-based stomatal conductance model (BMF model, Buckley et al., 2003) to estimate absolute changes in leaf turgor pressures. Turgor pressure of guard cells and surrounding epidermal cells determine Combining the BMF model with the LPCP probes 108 largely the stomatal aperture (Franks et al., 2001; Buckley & Mott, 2002b), and in this sense the use of BMF model to estimate turgor pressure from stomatal conductance measurements seems highly appropriated. Our main objective was to assess whether the diurnal leaf turgor pressure changes derived from LPCP outputs agrees with the turgor pressure modeled by the BMF model. Additionally, the concomitant use of LPCP probes and the BMF model will be used to understand mechanisms of stomatal conductance regulation under water stress. Experiments were carried out in a hedgerow olive tree orchard under two water treatments. Measurements were conducted at different leaf locations within the canopy, aiming to explore the behavior of stomata and main physiological variables determining the plant water status under different microenvironments. Material and methods Experimental design The experiment was conducted in 2012 in a commercial hedgerow olive orchard (Olea europaea L., cv. Arbequina) near Seville, Spain (37º 15’ N, −5º 48’ W). The orchard was planted with 1667 tree ha−1 in 2007. An experiment on irrigation started in 2010 (Fernández et al., 2013). Preliminary data show that differences among leaves within individual canopies, not differences among individuals, give rise to most of the variation in stomatal behavior in our study species. As our goal was not to describe differences among treatments, but to explain the physiological basis in the response of LPCP probes observed in a wide range of values, we decided to replicate locations within the canopy rather than tree individuals. Therefore, one tree per treatment in two of the four considered treatments was selected to carry out the study. The irrigation treatments supplied were: Control (well watered, WW), with daily irrigation to replace 100 % of the maximum Chapter 4 109 potential crop evapotranspiration (ETc); and a regulated deficit irrigation treatment (water stressed, WS), aimed to supply a total of 30 % ETc, with changing irrigation intensities and frequency during the season. Details on both the calculation of ETc and actual water supplies to each treatment are given in Fernández et al. (2013). The irrigation system consisted of one drip line per tree row with a 2 L h−1 dripper every 0.5 m. The soil in the orchard (Arenic Albaqualf, USDA 2010) had a sandy loam top layer of 0.6 m and a sandy clay layer downwards (Fernández et al., 2013). Soil and weather conditions Volumetric soil water content (  v) was estimated from measurements with a Profile probe (Delta-T Devices Ltd, Cambridge, UK) installed at 0.1 m from a dripper of the WS treatment. The probe was connected to a CR1000 Campbell datalogger (Campbell Scientific Ltd., Shepshed, UK) for collecting records every 10 min. In the WW treatment, where daily irrigation was supplied, we used a Profile probe to measure  v once or twice per week, in two access tubes (at 0.1 and 0.4 from a dripper). In all cases  v was measured at 0.1, 0.2, 0.3, 0.4, 0.6 and 1.0 m depths. We averaged  v down to 0.4 m depth only, since Diaz-Espejo et al. (2012) did not find roots in the orchard below 0.45 m. This was confirmed in this study by constant  v at 0.6 and 1.0 m depth along the studied period. The Profile probe was calibrated in situ by Fernández et al. (2011b). Weather variables were monitored by a Campbell weather station (Campbell Scientific Ltd., Shepshed, UK) installed in the centre of the experimental area. The meteorological sensors were installed at 3 m above the trees and average values of air temperature (Ta) and air vapour pressure deficit (VPD) were recorded every 30 min. Photosynthetic photon flux densities (PPFD) were monitored at the same time-step than stomatal conductance measurements (see below) with the photosynthetically active Combining the BMF model with the LPCP probes 116 Buckley et al. (2003) showed that  could be interpreted as the ratio of the specific rates of active ion uptake and passive ion efflux in guard cells. Because abscisic acid (ABA) affects guard cells by stimulating passive efflux (Hetherington, 2001), one would expect  – and therefore the parameter n in the modified model – to decline as ABA concentration increases. Equation (4.3) is a simplified form of the BMF model based on the following assumptions: the osmotic gradient from guard to epidermal cells, not to the apoplast, is the target for active regulation; the resistance from epidermal to guard cells is negligible compared to the resistance from the soil to the epidermis; and epidermal and bulk leaf osmotic pressure are similar. One of the constraints embedded in the BMF model is the ‘hydroactive feedback hypothesis’ (Buckley, 2005;  g =  e +  Pe, where  g is the guard cell osmotic pressure,  e is the epidermal cell osmotic pressure and Pe is the epidermal cell turgor pressure) which can be written as: (4.6)        leafs g, solving for  : (4.7) leaf s leaf sΨ na g Ψ g   . Equation (4.7) was used in the present work to estimate  with concurrent measurements of gs,  leaf and ci and modeled na on each time step. Parameters of photosynthetic capacity (Vc,max, Jmax and gm, Table 4.1) and diurnal measurements of ci were used to calculate  , embedded in the parameter na, meanwhile  were parameterize from fitting the BMF model to gs data. The maximum  value obtained in each day was selected as input Chapter 4 117 in the BMF model. See Results section for further explanations on why  measured was not used. Variable hydraulic conductance (Kvar) was obtained by fitting gs data to the BMF model at single measuring time solving Equation (4.3) as: (4.8) na g Ψ g Ks s s var VPD    . The modeled leaf turgor pressure (Pmodel) was derived from the standard expression of plant water relation: Pmodel =  l,model +  , where  l,model is the modeled leaf water potential estimated as: (4.9)       s models, modell, VPD   K g, and  is the maximum value estimated as indicated above. Statistics Linear mixed models were used to analyze effects of irrigation treatment (as a fixed factor) on  leaf, gs,  , K, n and  d. Leaf within position into the tree canopy was used as random factor when necessary to describe our experimental design. Homogeneity of variance and normality were tested and transformations were applied when needed. Models were fitted by restricted maximum likelihood (REML) in R (package ‘nlme R’; Pinheiro et al., 2012). When significant differences were obtained by ANOVA ( = 0.05), multiple comparisons were conducted for groups of more than two levels (package ‘multcomp’; Hothorn et al., 2008). Linear regression analyses for P’p and  leaf relationships were made by using SigmaPlot 11.0 (Systat Software, Inc., California, USA). Combining the BMF model with the LPCP probes 118 Results During the progress of the experiment, the WW treatment presented  v values around 0.25 m3 m−3 (Fig. 4.1), which in the soil of this study means soil capacity. The lower irrigation dose in June and the weekly irrigation events in August were reflected in the  v pattern in the WS treatment. Maximum air temperature (Ta) over 35°C and maximum evaporative demand (VPD) over 5 kPa were observed in the two daily cycle measurements. Despite of the high VPD on June 25, sky was partially cloudy, and maximum PPFD did not exceed 600 mol m−2 s−1 in the eastern sunny locations (Fig. 4.1E, F). Meanwhile, a maximum up to 800 mol m−2 s−1 was recorded in the west sunny location. Clear skies on August 3 allowed for a maximum PPFD between 1200-1400 mol m−2 s−1. As expected, the irradiance peaked at a different time of the day in east and west sides of the canopy (Fig. 4.1G, H). In both June and August, PPFD inside the canopy was up to 10-fold lower than in the outer canopy. Chapter 4 119 Fig. 4.1. Time courses of (A, B) volumetric soil water content (  v) measured on both well watered (WW) and water stressed (WS) treatments, (C, D) air temperature (Ta), air vapour pressure deficit (VPD) and (E, F, G, H) photosynthetically active photon flux density (PPFD) along the experiment. Panels (E) and (G) correspond to WW and (F) and (H) to WS treatments. Shaded areas indicate the two experimental dates, zoomed-in in the case of PPFD measurements. GMT = Greenwich Mean Time. On June 25 the differences in  leaf between treatments and locations were small and only significant in the afternoon. But, the effect of different VPD (kPa) 0 1 2 3 4 5 6 7 8 Jun 23 Jun 27Jun 25 Aug 1 Aug 3 Aug 5 Jun 29 Aug 7  v (m3 m-3) 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 WW WS A B C D E F Ta (ºC) 10 15 20 25 30 35 40 June 25 August 3 Time (h, GMT) PPFD (mol m-2 s-1) 0 200 400 600 800 1000 1200 1400 1600 East Shade West G H 07:00 12:00 17:00 12:00 17:00 12:00 17:00 12:00 17:00 Combining the BMF model with the LPCP probes 120 water supply was clearly observed on August 3 with a significant drop in both predawn and minimum  leaf in WS leaves (Fig. 4.2A, B). Both sunny and shaded leaves presented a similar decline in  leaf, which reached values as low as −4.2 MPa, 3 MPa lower than the WW leaves. Accordingly, these severe water stressed conditions led to a strong stomatal closure (Fig. 4.2D). Clearly different gs were observed between sunny and shaded leaves in both treatments, disappearing in the WS leaves once water stress was established in August (Fig. 4.2C, D). Furthermore, shaded leaves showed a more constant behavior along the studied period, and the main response to water stress came from sunny exposed leaves. Fig. 4.2. Leaf water potential (  leaf, −MPa) and stomatal conductance (gs, mol H2O m−2 s−1) measured in leaves from different locations within the canopy for the two water treatments (WW and WS) and on the two experimental dates. Error bars show standard errors for n = 3 (  leaf) and n = 4 (gs). Different letters indicate significant differences 05:00 09:00 13:00 17:00 Time (h, GMT) 05:00 09:00 13:00 17:00 gs (mol H2O m-2 s-1) 0.00 0.03 0.06 0.09 0.12 0.15 0.18 August 3June 25  leaf (MPa) -5.0 -4.5 -4.0 -3.5 -3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 WW east WW shade WW west WS east WS shade a ac ac b bc a b a c bc a a a b ba a a b ba a a b ba a a b b a a a b b a a a b b a a a b b ab b a c c a a a b b a a a b b a a a b b ab b a c ca a a b b a a a b b b b a c cb b a b bb b a b bb b a b b b c a d cd a b a c bc a b a b b a b a b b b c a d c b b a b b AB C D ab ac a c cb Chapter 4 121 (Multiple comparisons on significant effects from linear mixed model, P < 0.05). When no letters were presented, non significant differences were found. GMT = Greenwich Mean Time. The output of LPCP probes, although proportional to turgor pressure and  leaf, is dependent on clamping pressure and elastic properties of material clamped. This makes difficult averaging replicates. An example of the three probes installed in two locations in this study is shown in Fig. 4.3. WS leaves were chosen because of the wider range of  leaf shown, as well as both sunny and shaded locations. It can be observed that each probe varied in its offset, which motivated that probe readings were normalized following Equation (4.2). Measurements of  leaf are plotted on top of them for comparison (open symbols in Fig. 4.3). Significant negative linear relationships were obtained between P’p and  leaf (Fig. 4.4), except in some cases where the correlation showed a hysteretic behavior. This phenomenon was clearly evident in both eastern and shaded leaves of the WS leaves on August 3 (Fig. 4.4B, D), but also slightly presented in eastern leaves in June (Fig. 4.4A). Although a linear relationship between P’p and  leaf was observed in both treatments and dates, in the WS leaves the correlations were shifted to much lower values of  leaf in August. The WW leaves also presented a slight shift towards the same direction (note different range of X-axis in Fig. 4.4 between left and right panels). Combining the BMF model with the LPCP probes 122 Fig. 4. 3. Leaf patch clamp pressure (LPCP) probe actual recordings during August 3 in three eastern and shaded leaves of the WS tree (lines). Simultaneous leaf water potential (  leaf, −MPa) measurements in the same canopy locations are also presented (open symbols). Error bars show standard errors for n = 3. GMT = Greenwich Mean Time. August 3 Time (h, GMT) 00:00 05:00 10:00 15:00 20:00 Pp (kPa) 0 20 40 60 80 100 120 140 00:00 05:00 10:00 15:00 20:00  leaf (MPa) -4 -3 -2 -1 0 East Shade Chapter 4 123 Fig. 4. 4. Relationships between the normalized output leaf patch pressure ( P’ p , %) and the leaf water potential (  leaf, −MPa) measured on the two experimental dates in different leaf locations within the canopy and for the two water treatments (WW and WS). Error bars show standard errors for n = 3. Gray arrows indicate the diurnal evolution of the measurements when hysteresis was observed. The stomatal conductance model (BMF model) was used to estimate the diurnal progress of leaf turgor pressure (Pmodel), and to relate it to the Pp dynamics. Initially, the model was incapable of reproducing simultaneously June 25 East 0.0 0.2 0.4 0.6 0.8 1.0 1.2 WW WS  leaf (MPa) -1.4-1.2-1.0-0.8-0.6-0.4-0.20.0 West 0.0 0.2 0.4 0.6 0.8 1.0 August 3 P'p (%) Shade 0.0 0.2 0.4 0.6 0.8 1.0 -5-4-3-2-10 A B C D E F Combining the BMF model with the LPCP probes 124 the diurnal behavior of both gs and  leaf. This is a process-based model of stomatal conductance that, among several assumptions, uses a constant  during the day. Under the conditions of severe water stress experienced by the WS leaves, osmotic pressures measured at dawn (  d; Table 4.2) were too low to reproduce the low  leaf measured in the afternoon (Fig. 4.2B). To solve that discrepancy, we used a maximum estimated  value based on  leaf as explained in M&M. Diurnal change of  has been reported by several authors (Girma & Krieg, 1992; Patakas & Noitsakis, 1999; Hummel et al., 2010; Himmelsbach et al., 2012), reaching increments of 0.5 - 1.4 MPa in olive to tolerate water stress periods maintaining photosynthetic activity (Dichio et al., 2006). The BMF model satisfactorily captured the diurnal evolution observed in gs in both treatments (Fig. 4.5) and produced parameters with full physiological meaning (Fig. 4.6). Accounting that  were estimated values from  leaf, a large osmotic adjustment was observed in both sunny and shaded leaves in WS from June to August (Fig. 4.6A). Soil-to-leaf hydraulic conductance (K) was quite constant for WW sunny leaves in both June and August, meanwhile in the WS leaves K decreased in August. Shaded leaves tended to have lower K than sunny ones, independently of water treatment, being larger the difference in August for WW, but disappearing in WS. However, the non-hydraulic term n did not have a distinguishable pattern. On June 25, a high variability was observed probably due to the cloudy weather conditions and due to the close relation of this parameter with PPFD through the embedded parameter  . Despite of that, n mostly declined in August in both sunny and shaded leaves reaching lower values in the WS leaves. Although the goodness of fit of the BMF model presented in Fig. 4.5 (that assumed all the estimated parameters constant during the day) can be considered as good, still the model was not able to interpret some points at the onset and end of the day (e.g. Fig. 4.5A). To account for these discrepancies, we evaluated a dynamic diurnal variation in K (Kvar) as a possible physiological mechanism not considered in the model at its current version. In order to achieve this, Kvar was allowed to Chapter 4 125 change to fit gs data to the BMF model output at very single measuring time. The new modeled gs is not shown since simply matches perfectly with gs data, but the diurnal evolution of Kvar modeled is shown in Fig. 4.7. Results suggested that K changed during the day in both treatments. Additionally, Kvar was plotted as a function of  leaf (Fig. 4.8). Leaf turgor pressures modeled by the BMF model under a dynamic K were plotted against the normalized values of Pp (Fig. 4.9). Both variables were correlated and followed a power function. This power function was mainly observed in August. Table 4.2. Osmotic pressures measured at dawn (  d) on the two daily cycle measurements. Numbers between brackets show standard errors (n = 3). Different letters indicate significant differences (Multiple comparisons on significant effects from linear mixed model, P < 0.05). Experimental period Treatments and positions  d MPa June WW East 1.80 (0.08) cb Shade 1.50 (0.03) b West 1.91 (0.11) cb WS East 2.48 (0.16) a Shade 2.18 (0.09) ac August WW East 1.84 (0.09) cb Shade 2.00 (0.09) ac West 2.05 (0.15) ac WS East 2.28 (0.05) ac Shade 2.23 (0.16) ac Appendix 228 (II.16)             wgggKgagngg Δ ssssssss                , where  gs(n) = (∂gs/∂n)  n and so forth. Finally, the proportional (or percent) contribution of each parameter can be estimated by dividing through by the sum of these changes. For example, the percent contribution due to nonhydraulic factors other than ATP (represented by n) is −100·  gs(n)/(  gs(n) +  gs(a) +  gs(K) +  gs(  s) +  gs(  ) +  gs(w)). (The negative sign ensures that the direction of each change is retained in the resulting percent changes; for example, if both  gs(K) and the overall change in gs are negative, as they are in the example studied in the main text, the negative effect on gs of a decline in K will be represented as a negative number.) The partial derivatives in Equation (II.15) are easily computed by differentiating Equation (II.11) to give (II.17)   a wnaK K n g             s 2 s Δ, (II.18)   n wnaK K a g             s 2 s Δ, (II.19)   w wnaK na K gΔ Δs 2 s            , (II.20) wnaK naK g Δ s s      , (II.21) w na K naKg Δ s     , and (II.22)   K wnaK na w g             s 2 s ΔΔ . 229 To apply these expressions to (II.15) and (II.16) in order to examine changes in gs between two days in the WS treatment (21 and 31 Aug) at a given hour of the day, we used values of each parameter or variable in (II.17)-(II.22) corresponding to that time of day, but averaged between the two days (e.g., calculations for Equation II.17 applied at 1300h would use w = 0.5(w(1300h, 21 Aug) + w(1300h, 31 Aug)). Temperature response of photosynthetic model parameters Several parameters of the Farquhar et al. (1980) model of photosynthesis used in the present work and shown in Equations (II.8) and (II.9) have a high dependence on temperature. We used the values proposed by Bernacchi et al. (2002) for Kc, Ko and  * obtained in tobacco in vivo in a ccbasis. (II.23) )/Δ(exp ka RTHcParameter   where c is the scaling constant, Ha the energy of activation, and Tk leaf temperature in ºK. The specific temperature response of Vm, J and gm in almond was obtained from Egea et al. (2011b) (II.24)            k dk ka 25 ΔΔ exp1 )/Δ(exp RT HST RTHc ParameterParameter where Parameter25 is the value of the parameter at 25 ºC, Hd the energy deactivation and S an entropy term. Values of the parameters in (II.27) and (II.28) used in this chapter are given in Extra Table 3.1. Appendix 230 Extra Table 3.1. Parameter values used in this chapter for responses of photosynthetic parameters to temperature. c  Ha  Hd  S K c 38.28 80.99 K o 14.68 23.72  * 13.49 24.46 V m 31.57 78.05 155.20 0.5 J 14.03 34.75 189.79 0.6 g m 19.94 49.43 431.12 1.37 Appendix III. Chapter 6. Other measurements For each plot profiles of  v were measured 1-2 times per week using a Profile probe (Delta-T Devices Ltd, Cambridge, UK). The Profile probe was calibrated in situ, by comparing the soil permittivity (έ) values derived from the Profile probe readings with  v values measured with time domain reflectometry (TDR) probes. The resulting calibration curve was: (έ)0.5 = 2.177 + 6.66  v, r2 = 0.72. At the plots which were subjected to the RDI treatments we installed two access tubes per plot, in a distance of 0.5 m from the tree trunk and of 0.01 m and 0.04 m, respectively, from the dripper. In the Control plot we placed six access tubes, three at a distance of 0.01 m and three at a distance of 0.04 m from the dripper. Measurements in each access tube were made at 0.1, 0.2, 0.3, 0.4, 0.6 and 1.0 m depths. The 231  v values in the rootzone, i.e. down to 0.6 m, were used to derive (according to Granier, 1987) a depth equivalent of water expressed as the level of relative extractable water (REW) (Extra Figure 6.1C). The time course of the water status of the trees was monitored by measuring the leaf water potential at predawn (  pd) and the midday stem water potential (  stem), once every two weeks during the whole irrigation season. Measurements were made with a Scholander-type pressure chamber (PMS Instrument Company, Albany, Oregon, USA). For the RDI treatments we sampled one leaf per tree from two representative trees per plot (n = 8). In the Control plot we sampled two leaves per tree from four trees in order to have the same number of replicates. For  pd we sampled the 4th or 5th leaf apart from the apex of peripheral twigs at about 1.5 - 1.9 m height above the ground. For  stem we sampled leaves from the inner part of the canopy. The leaves were wrapped in aluminium foil ca. 2 h before the measurements. Comparative measurements of Pp and leaf water potential (  leaf) were performed on June 23 and 24 as well as on September 9, i.e. before and after the period of severe water restrictions suffered by the RDI trees at midsummer. At these days, we measured  leaf of each treatment every hour taking at least 8 replicates from dawn to sunset. Leaves were sampled as described for  pd following the path of the sun. Thirty minute average values of the main meteorological variables were recorded by a Campbell weather station (Campbell Scientific Ltd., Shepshed, UK) located in the centre of the area covered by the experimental plots. We used a pole for installing all meteorological sensors between 2 m and 3 m height above the canopies. 233 References Ache P, Bauer H, Kollist H, Al-Rasheid KAS, Lautner S, Hartung W, Hedrich R. 2010. Stomatal action directly feeds back on leaf turgor: new insights into the regulation of the plant water status from non-invasive pressure probe measurements. The Plant Journal 62: 1072 - 1082. AEA. 2011. Anuario de Estadística Agraria. Ministerio de Agricultura, Alimentación y Medio Ambiente. España. Alder NN, Sperry JS, Pockman WT. 1996. Root and stem xylem embolism stomatal conductance, and leaf turgor in Acer grandidentatum populations along a soil moisture gradient. Oecologia 105: 293 - 301. Allen RG, Pereira LS, Raes D, Smith M. 1998. Crop evapotranspiration. Guidelines for computing crop water requirements. Irrigation and Drainage paper 56. FAO, Rome, Italy. Angeles G, et al. 2004. The Cohesion-Tension Theory. Letters. New Phytologist 163: 451 - 452. Aranda I, Forner A, Cuesta B, Valladares F. 2012. Species-specific water use by forest tree species: from the tree to the stand. Agricultural Water Management 114: 67 - 77. Ball JT, Woodrow IE, Berry JA. 1987. A model predicting stomatal conductance and its contribution to the control of photosynthesis under different environmental conditions. In: Progress in Photosynthesis Research (ed. J. Biggens), pp. 221 - 224. Martinus Nijhoff Publishers, The Netherlands. Barranco D. 2001. Variedades y patrones. En: El Cultivo del olivo. 4ª Edición, pp. 63 - 88. Barranco D., Fernández-Escobar R. & Rallo L. Ed. Mundi-Prensa, Junta de Andalucía. Barros PM, Gonçalves N, Saibo NJM, Oliveira MM. 2012. Cold acclimation and floral development in almond bud break: insights into the regulatory pathways. Journal of Experimental Botany 63: 4585 - 4596. 234 Bartlett MK, Scoffoni C, Ardy R, Zhang Y, Sun S, Cao K, Sack L. 2012. Rapid determination of comparative drought tolerance traits: using an osmometer to predict turgor loss point. Methods in Ecology and Evolution 3: 880 - 888. Bauer H, Ache P, Lautner S, Fromm J, Hartung W, Al-Rasheid K a S, Sonnewald S, Sonnewald U, Kneitz S, Lachmann N, Mendel R, Bittner F, Hetherington AM, Hedrich R. 2013. The stomatal response to reduced relative humidity requires guard cell-autonomous ABA synthesis. Current Biology 23: 53 - 57. Beikircher B, Mayr S. 2009. Intraspecific differences in drought tolerance and acclimation in hydraulics of Ligustrum vulgare and Viburnum lantana. Tree Physiology 29: 765 - 775. Ben-Gal A, Kool D, Agam N, van Halsema GE, Yermiyahu U, Yafe A, Presnov E, Erel R, Majdop A, Zipori I. 2010. Whole-tree water balance and indicators for short-term drought stress in non-bearing “Barnea” olives. Agricultural Water Management 98: 124 - 133. Bernacchi CJ, Portis AR, Nakano H, von Caemmerer S, Long SP. 2002. Temperature response of mesophyll conductance. Implications for the determination of Rubisco enzyme kinetics and for limitations to photosynthesis in vivo. Plant Physiology 130: 1992 - 1998. Blank K, Lamersdorf N, Dohrenbusch A, Murach D. 1995. Response of a Norway spruce forest ecosystem to drought ⁄ rewetting experiments at Solling, Germany. Water, Air and Soil Pollution 85: 1251 - 1256. Boyer JS. 1967. Leaf Water Potentials Measured with a Pressure Chamber. Plant Physiology 42: 133 - 137. Boyer JS. 1982. Plant productivity and environment. Science 218: 443 - 448. Boyer JS. 1996. Advances in drought tolerance in plants. Advances in Agronomy 56: 187 - 218. Bramley H, Ehrenberger W, Zimmermann U, Palta JA, Rüger S, Siddique KHM. 2013. Non-invasive pressure probes magnetically clamped to leaves to monitor the water status of wheat. Plant and Soil 369: 257 - 268. 235 Bramley H, Turner NC, Turner DW, Tyerman SD. 2007. Comparison between gradient-dependent hydraulic conductivities of roots using the root pressure probe: the role of pressure propagations and implications for the relative roles of parallel radial pathways. Plant, Cell & Environment 30: 861 - 874. Brodribb TJ, Cochard H. 2009. Hydraulic failure defines the recovery and point of death in water-stressed conifers. Plant Physiology 149: 575 - 584. Brodribb TJ, Holbrook NM. 2003. Stomatal Closure during Leaf Dehydration, Correlation with Other Leaf Physiological Traits. Plant Physiology 132: 2166 - 2173. Brodribb TJ, Holbrook NM. 2004. Diurnal depression of leaf hydraulic conductance in a tropical tree species. Plant, Cell & Environment 27: 820 - 827. Brodribb TJ, McAdam SAM. 2011. Passive origins of stomatal control in vascular plants. Science 331: 582 - 585. Brodribb TJ, McAdam SAM. 2013. Abscisic Acid mediates a divergence in the drought response of two conifers. Plant Physiology 162: 1370 - 1377. Buckley TN, Mott KA, Farquhar GD. 2003. A hydromechanical and biochemical model of stomatal conductance. Plant, Cell & Environment 26: 1767 - 1785. Buckley TN, Mott KA. 2002a. Dynamics of stomatal water relations during the humidity response: implications of two hypothetical mechanisms. Plant, Cell & Environment 25: 407 - 419. Buckley TN, Mott KA. 2002b. Stomatal water relations and the control of hydraulic supply and demand. Progress in Botany 63: 309 - 325. Buckley TN, Mott KA. 2013. Modelling stomatal conductance in response to environmental factors. Plant, Cell & Environment 36: 1691 - 1699. Buckley TN. 2005. The control of stomata by water balance. The New Phytologist 168: 275 - 292. Burgess SSO, Adams MA, Bleby TM. 2000. Measurement of sap flow in roots of woody plants: a commentary. Tree Physiology 20: 909 - 913. 236 Burquez A. 1987. Leaf thickness and water deficits in plants: a tool for field studies. Journal of Experimental Botany 38: 109 - 114. Callister AN, Arndt SK, Adams MA. 2006. Comparison of four methods for measuring osmotic potential of tree leaves. Physiologia Plantarum 127: 383 - 392. Campbell GS. 1985. Soil physics with BASIC - Transport models for soil-plant systems. In: Developments in Soil Science 14. Elsevier, New York. Cardon ZG, Mott KA, Berry JA. 1994. Dynamics of patchy stomatal movements, and their contribution to steady-state and oscillating stomatal conductance calculated using gas-exchange techniques. Plant, Cell & Environment 17: 995 - 1007. Carr MKV. 2013. The water relations and irrigation requirements of olive (Olea europaea L.): a review. Experimental Agriculture 49: 597 - 639. Cermák J, Kucera J, Bauerle WL, Phillips N, Hinckley TM. 2007. Tree water storage and its diurnal dynamics related to sap flow and changes in stem volume in old-growth Douglas-fir trees. Tree Physiology 27: 181 - 198. Chalmers DJ, Mitchell PD, van Heek L. 1981. Control of peach tree growth and productivity by regulated water supply, tree density and summer pruning. Journal of the American Society for Horticultural Science 106: 307 - 312. Chaves MM, Pereira JS, Maroco J, Rodrigues ML, Ricardo CPP, Osório ML, Carvalho I, Faria T, Pinheiro C. 2002. How plants cope with water stress in the field? Photosynthesis and growth. Annals of Botany 89: 907 - 916. Choat B, Drayton WM, Brodersen C, Matthews MA, Shackel KA, Wada H, McElrone AJ. 2010. Measurement of vulnerability to water stress-induced cavitation in grapevine: a comparison of four techniques applied to a long-vesseled species. Plant, Cell & Environment 33: 1502 - 1512. Christmann A, Grill E, Huang J. 2013. Hydraulic signals in long-distance signaling. Current Opinion in Plant Biology 16: 293 - 300. Christmann A, Hoffmann T, Teplova I, Grill E, Müller A. 2005. Generation of active pools of abscisic acid revealed by in vivo imaging of water-stressed Arabidopsis. Plant Physiology 137: 209 - 219. 237 Christmann A, Weiler EW, Steudle E, Grill E. 2007. A hydraulic signal in rootto-shoot signalling of water shortage. The Plant Journal 52: 167 - 174. Cifre J, Bota J, Escalona JM, Medrano H, Flexas J. 2005. Physiological tools for irrigation scheduling in grapevine (Vitis vinifera L.). Agriculture, Ecosystems & Environment 106: 159 - 170. Clark J, Gibbs RD. 1957. Studies in tree physiology. IV. Further investigations of seasonal changes in moisture content of certain Canadian forest trees. Canadian Journal of Botany 35: 219 - 253. Cochard H, Barigah ST, Kleinhentz M, Eshel A. 2008. Is xylem cavitation resistance a relevant criterion for screening drought resistance among Prunus species? Journal of Plant Physiology 165: 976 - 982. Comstock JP, Mencuccini M. 1998. Control of stomatal conductance by leaf water potential in Hymenoclea salsola (T. & G.), a desert subshrub. Plant, Cell & Environment 21: 1029 - 1038. Conejero W, Alarcón JJ, García-Orellana Y, Nicolás E, Torrecillas A. 2007. Evaluation of sap flow and trunk diameter sensors for irrigation scheduling in early maturing peach trees. Tree Physiology 27: 1753 - 1759. Connor DJ. 2005. Adaptation of olive (Olea europaea L.) to water-limited environments. Australian Journal of Agricultural Research 56: 1181 - 1189. Cornwell WK, Bhaskar R, Sack L, Cordell S, Lunch CK. 2007. Adjustment of structure and function of Hawaiian Metrosideros polymorpha at high vs low precipitation. Functional Ecology 21: 1063 - 1071. Cowan IR, Farquhar GD. 1977. Stomatal function in relation to leaf metabolism and environment: stomatal function in the regulation of gas exchange. In Integration of Activity in the Higher Plant (ed. D.H. Jennings), pp. 471 - 505. Cambridge University Press, Cambridge, UK. Cuevas MV, Martín-Palomo MJ, Diaz-Espejo A, Torres-Ruiz JM, Rodriguez-Dominguez CM, Perez-Martin A, Pino-Mejías R, Fernández JE. 2013. Assessing water stress in a hedgerow olive orchard from sap flow and trunk diameter measurements. Irrigation Science 31: 729 - 746. 244 Giorio P, Sorrentino G, D’Andria R. 1999. Stomatal behaviour, leaf water status and photosynthetic response in field-grown olive trees under water deficit. Environmental and Experimental Botany 42: 95 - 104. Girma FS, Krieg DR. 1992. Osmotic Adjustment in Sorghum: I. Mechanisms of Diurnal Osmotic Potential Changes. Plant Physiology 99: 577 - 582. Girona J, Mata M, Marsal J. 2005. Regulated deficit irrigation during the kernelfilling period and optimal irrigation rates in almond. Agricultural Water Management 75: 152 - 167. Goldhamer D, Fereres E. 2001. Irrigation scheduling protocols using continuously recorded trunk diameter measurements. Irrigation Science 20: 115 - 125. Goldhamer DA, Viveros M, Salinas M. 2006. Regulated deficit irrigation in almonds: effects of variations in applied water and stress timing on yield and yield components. Irrigation Science 24: 101 - 114. Goldhamer DA, Viveros M. 2000. Effects of preharvest irrigation cutoff durations and postharvest water deprivation on almond tree performance. Irrigation Science 19: 125 - 131. Gómez‐Cadenas A, Pozo OJ, García‐Augustín P, Sancho JV. 2002. Direct analysis of abscisic acid in crude plant extracts by liquid chromatographyelectrospray/tandem mass spectrometry. Phytochemical Analysis 13: 228 - 234. Gómez-del-Campo M. 2013. Summer deficit-irrigation strategies in a hedgerow olive orchard cv. “Arbequina”: effect on fruit characteristics and yield. Irrigation Science 31: 259 - 269. González-Altozano JR, Castel P. 2003. Riego deficitario controlado en “Clementina de Nules”. I. Efectos sobre la producción y la calidad de la fruta. Spanish Journal of Agricultural Research 1: 81 - 92. Graham JB, Aguilar NM, Dudley R, Gans C. 1995. Implications of the late Palaeozoic oxygen pulse for physiology and evolution. Nature 375: 117 - 120. Granier A. 1987. Evaluation of transpiration in a Douglas-fir stand by means of sap flow measurements. Tree Physiology 3: 309 - 320. 245 Grant OM, Davies MJ, James CM, Johnson AW, Leinonen I, Simpson DW. 2012. Thermal imaging and carbon isotope composition indicate variation amongst strawberry (Fragaria × ananassa) cultivars in stomatal conductance and water use efficiency. Environmental and Experimental Botany 76: 7 - 15. Grassi G, Magnani F. 2005. Stomatal, mesophyll conductance and biochemical limitations to photosynthesis as affected by drought and leaf ontogeny in ash and oak trees. Plant, Cell & Environment 28: 834 - 849. Green S, Clothier B, Jardine B. 2003. Theory and Practical Application of Heat Pulse to Measure Sap Flow. Agronomy Journal 95: 1371 - 1379. Gurovich L, Hermosilla P. 2009. Electric signaling in fruit trees in response to water applications and light-darkness conditions. Journal of Plant Physiology 166: 290 - 300. Gutschick VP, Simonneau T. 2002. Modelling stomatal conductance of fieldgrown sunflower under varying soil water content and leaf environment: comparison of three models of stomatal response to leaf environment and coupling with an abscisic acid-based model of stomatal response to soil drying. Plant, Cell & Environment 25: 1423 - 1434. Guyot G, Scoffoni C, Sack L. 2012. Combined impacts of irradiance and dehydration on leaf hydraulic conductance: insights into vulnerability and stomatal control. Plant, Cell & Environment 35: 857 - 871. Hacke UG, Sperry JS, Ewers BE, Ellsworth DS, Schafer KVR, Oren R. 2000. Influence of soil porosity on water use in Pinus taeda. Oecologia 124: 495 - 505. Haefner JW, Buckley TN, Mott KA. 1997. A spatially explicit model of patchy stomatal responses to humidity. Plant, Cell & Environment 20: 1087 - 1097. Hetherington AM, Woodward FI. 2003. The role of stomata in sensing and driving environmental change. Nature 424: 901 - 908. Hetherington AM. 2001. Guard Cell Signaling. Cell 107: 711 - 714. Higgins SS, Larsen FE, Bendel RB, Radamaker GK, Bassman JH, Bidlake WR, Alwir A. 1992. Comparative gas-exchange characteristics of potted, 246 glasshouse-grown almond, apple, fig, grape, olive, peach and Asian pear. Scientia Horticulturae 52: 313 - 329. Himmelsbach W, Treviño-Garza EJ, González-Rodríguez H, GonzálezTagle MA, Gómez Meza MV, Aguirre Calderón OA, Eduardo Estrada Castillón A, Mitlöhner R. 2011. Acclimatation of three co-occurring tree species to water stress and their role as site indicators in mixed pine-oak forests in the Sierra Madre Oriental, Mexico. European Journal of Forest Research 131: 355 - 367. Hinckley TM, Lassoie JP, Running SW. 1978. Temporal and spatial variations in the water status of forest trees. Forest Science 24: 1 - 72. Holbrook NM, Shashidhar VR, James RA, Munns R. 2002. Stomatal control in tomato with ABA-deficient roots: response of grafted plants to soil drying. Journal of Experimental Botany 53: 1503 - 1514. Holbrook NM, Zwieniecki MA. 2005. Vascular Transport in Plants. Academic Press, Amsterdam, 564 pp. Hothorn T, Bretz F, Westfall P. 2008. Simultaneous inference in general parametric models. Biometrical Journal 50: 346 - 363. Hu FB. 2003. The Mediterranean diet and mortality-olive oil and beyond. The New England Journal of Medicine 348: 2595 - 2596. Hummel I, Pantin F, Sulpice R, Piques M, Rolland G, Dauzat M, Christophe A, Pervent M, Bouteillé M, Stitt M, Gibon Y, Muller B. 2010. Arabidopsis plants acclimate to water deficit at low cost through changes of carbon usage: an integrated perspective using growth, metabolite, enzyme, and gene expression analysis. Plant Physiology 154: 357 - 372. Hutmacher RB, Nightingale HI, Rolston DE, Biggar JW, Dale F, Vail SS, Peters D. 1994. Growth and yield responses of almond (Prunus amygdalus) to trickle irrigation. Irrigation Science 14: 117 - 126. Iniesta F, Testi L, Orgaz F, Villalobos FJ. 2009. The effects of regulated and continuous deficit irrigation on the water use, growth and yield of olive trees. European Journal of Agronomy 30: 258 - 265. 247 Intrigliolo DS, Castel JR. 2006a. Performance of various water stress indicators for prediction of fruit size response to deficit irrigation in plum. Agricultural Water Management 83: 173 - 180. Intrigliolo DS, Castel JR. 2006b. Usefulness of diurnal trunk shrinkage as a water stress indicator in plum trees. Tree Physiology 26: 303 - 311. Intrigliolo DS, Castel JR. 2010. Response of plum trees to deficit irrigation under two crop levels: tree growth, yield and fruit quality. Irrigation Science 28: 525 - 534. IOC. International Olive Council, http://www.internationaloliveoil.org. IPCC. 2013. Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, New York. Irvine J, Perks MP, Magnani F, Grace J. 1998. The response of Pinus sylvestris to drought: stomatal control of transpiration and hydraulic conductance. Tree Physiology 18: 393 - 402. ISHS. Working Group on Sap Flow, http://www.wgsapflow.com Jacobsen S-E, Jensen CR, Liu F. 2012. Improving crop production in the arid Mediterranean climate. Field Crops Research 128: 34 - 47. Jarvis AJ, Davies WJ. 1998. The coupled response of stomatal conductance to photosynthesis and transpiration. Journal of Experimental Botany 49: 399 - 406. Jarvis PG, McNaughton KG. 1986. Stomatal control of transpiration: scaling up from leaf to region. Advances in Ecological Research 15: 1 - 49. Jarvis PG. 1976. The interpretation of the variations in leaf water potential and stomatal conductance found in canopies in the field. Philosophical Transactions of the Royal Society of London. Series B 273: 593 - 610. Jiang M, Zhang J. 2001. Effect of abscisic acid on active oxygen species, antioxidative defense system and oxidative damage in leaves of maize seedlings. Plant and Cell Physiology 42: 1265 - 1273. Johnson DM, Woodruff DR, McCulloh KA, Meinzer FC. 2009. Leaf hydraulic conductance, measured in situ, declines and recovers daily: leaf hydraulics, water 248 potential and stomatal conductance in four temperate and three tropical tree species. Tree Physiology 29: 879 - 887. Jones HG, Tardieu F. 1998. Modeling water relations of horticultural crops: A review. Scientia Horticulturae 74: 21 - 46. Jones HG 1985. Partitioning stomatal and non-stomatal limitations to photosynthesis. Plant, Cell & Environment 8: 95 - 104. Jones HG. 1990. Physiological aspects of the control of water status in horticultural crops. HortScience 25: 19 - 26. Jones HG. 1992. Stomata. In Plants and Microclimate: A Quantitative Approach to Environmental Plant Physiology (ed. H.G. Jones) pp. 131 - 162. Cambridge University Press, Cambridge, UK. Jones HG. 1999. Use of infrared thermometry for estimation of stomatal conductance as a possible aid to irrigation scheduling. Agricultural and Forest Meteorology 95: 139 - 149. Jones HG. 2002. Use of infrared thermography for monitoring stomatal closure in the field: application to grapevine. Journal of Experimental Botany 53: 2249 - 2260. Jones HG. 2004. Irrigation scheduling: advantages and pitfalls of plant-based methods. Journal of Experimental Botany 55: 2427 - 2436. Jones HG. 2007. Monitoring plant and soil water status: established and novel methods revisited and their relevance to studies of drought tolerance. Journal of Experimental Botany 58: 119 - 130. Kijne JW, Barker R, Molden DJ. 2003. Water Productivity in Agriculture: Limits and Opportunities for Improvement. CABI, IWMI, Wallingford, UK. Klein I, Esparza G, Weinbaum SA, DeJong TM. 2001. Effects of irrigation deprivation during the harvest period on leaf persistence and function in mature almond trees. Tree Physiology 21: 1063 - 1072. Klein M, Cheng G, Chung M, Tallman G. 1996. Effects of turgor potentials of epidermal cells neighbouring guard cells on stomatal opening in detached leaf epidermis and intact leaflets of Vicia faba L. (faba bean). Plant, Cell & Environment 19, 1399 - 1407. 249 Koch GW, Sillett SC, Jennings GM, Davis SD. 2004. The limits to tree height. Nature 428: 851 - 854. Kolb KJ, Sperry JS. 1999. Transport constraints on water use by the Great Basin shrub, Artemisia tridentata. Plant, Cell & Environment 22: 925 - 935. Kozlowski TT, Pallardy SG. 2002. Acclimation and Adaptive Responses of Woody Plants to Environmental Stresses. The Botanical Review 68: 270 - 334. Kramer PJ, Boyer JS. 1995. Water Relations of Plants and Soils. Academic Press, San Diego, CA, USA. Le Roux X, Grand S, Dreyer E, Daudet F-A. 1999. Parameterization and testing of a biochemically based photosynthesis model for walnut (Juglans regia) trees and seedlings. Tree Physiology 19: 481 - 492. Lee KM, Driever SM, Heuvelink E, Rüger S, Zimmermann U, de Gelder A, Marcelis LFM. 2012. Evaluation of diel patterns of relative changes in cell turgor of tomato plants using leaf patch clamp pressure probes. Physiologia Plantarum 146: 439 - 447. Leuning R. 1990. Modelling Stomatal Behaviour and Photosynthesis of Eucalyptus grandis. Australian Journal of Plant Physiology 17: 159 - 175. Leuning R. 1995. A critical appraisal of a combined stomatal-photosynthesis model for C3 plants. Plant, Cell & Environment 18: 339 - 355. Lo Gullo MA, Nardini A, Trifilò P, Salleo S. 2005. Diurnal and seasonal variations in leaf hydraulic conductance in evergreen and deciduous trees. Tree Physiology 25: 505 - 512. Lopez-Zamora I, Falcao N, Comerford NB, Barros NF. 2002. Root isotropy and an evaluation of a method for measuring root distribution in soil trenches. Forest Ecology and Management 166: 303 - 310. Maherali H, DeLucia EH. 2000. Xylem conductivity and vulnerability to cavitation of ponderosa pine growing in contrasting climates. Tree Physiology 20: 859 - 867. Marchi S, Tognetti R, Minnocci A, Borghi M, Sebastiani L. 2008. Variation in mesophyll anatomy and photosynthetic capacity during leaf development in a 250 deciduous mesophyte fruit tree (Prunus persica) and an evergreen sclerophyllous Mediterranean shrub (Olea europaea). Trees 22: 559 - 571. Marigo G, Peltier JP. 1996. Analysis of the diurnal change in osmotic potential in leaves of Fraxinus excelsior L. Journal of Experimental Botany 47: 763 - 769. Martínez-Vilalta J, Prat E, Oliveras I, Piñol J. 2002. Xylem hydraulic properties of roots and stems of nine Mediterranean woody species. Oecologia 133: 19 - 29. Matos MC, Rebelo E, Lauriano J, Semedo J, Marques N, Campos PS, Matos A, Vieira-Da-Silva J. 2004. CO2 Assimilation and Water Relations of Almond Tree (Prunus amygdalus Batsch) Cultivars Grown Under Field Conditions. Photosynthetica 42: 473 - 476. McAdam SAM, Brodribb TJ. 2012. Stomatal innovation and the rise of seed plants. Ecology Letters 15: 1 - 8. McBurney T. 1988. A temperature-controlled plant psychrometer. Plant and Soil, 109: 271 - 275. McCarthy MG. 1997. The effect of transient water deficit on berry development of cv. Shiraz (Vitis vinifera L.). Australian Journal of Grape and Wine Research 3: 2 - 8. McDowell N, Pockman WT, Allen CD, Breshears DD, Cobb N, Kolb T, Plaut J, Sperry JS, West A, Williams DG, et al. 2008. Mechanisms of plant survival and mortality during drought: why do some plants survive while others succumb to drought? The New phytologist 178: 719 - 739. Meinzer FC. 2002. Co-ordination of vapour and liquid phase water transport properties in plants. Plant, Cell & Environment 25: 265 - 274. Messinger SM, Buckley TN, Mott KA. 2006. Evidence for involvement of photosynthetic processes in the stomatal response to CO2. Plant Physiology 140: 771 - 778. Miranda JD, Padilla FM, Martínez-Vilalta J, Pugnaire FI. 2010. Woody species of a semi-arid community are only moderately resistant to cavitation. Functional Plant Biology 37: 828 - 839. 251 Mitchell PD, Chalmers DJ. 1982. The effects of reduced water supply on peach tree growth and yield. Journal of the American Society for Horticultural Science 107: 853 - 856. Möller M, Alchanatis V, Cohen Y, Meron M, Tsipris J, Ostrovsky V, Sprintsin M, Cohen S. 2007. Use of thermal and visible imagery for estimating crop water status of irrigated grapevine. Journal of Experimental Botany 58: 827 - 838. Monteith JL. 1995. A reinterpretation of stomatal responses to humidity. Plant, Cell & Environment 18: 357 - 364. Moreno F, Fernández JE, Clothier BE, Green SR. 1996. Transpiration and root water uptake by olive trees. Plant and Soil 184: 85 - 96. Moriana A, Orgaz F, Pastor M, Fereres E. 2003. Yield responses of a mature olive orchard to water deficits. Journal of the American Society for Horticultural Science 128: 425 - 431. Moriana A, Villalobos FJ, Fereres E. 2002. Stomatal and photosynthetic responses of olive (Olea europaea L.) leaves to water deficits. Plant, Cell & Environment 25: 395 - 405. Morison JIL, Baker NR, Mullineaux PM, Davies WJ. 2008. Improving water use in crop production. Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences 363: 639 - 658. Mott KA, Parkhurst DF. 1991. Stomatal responses to humidity in air and helox. Plant, Cell & Environment 14: 509 - 515. Mott KA. 1988. Do Stomata Respond to CO2 Concentrations Other than Intercellular? Plant Physiology 86: 200 - 203. Münch E. 1930. Die Stoffbewegungen in den Pflanzen. Gustav Fischer, Jena, Germany. Murphy R, Ortega JKE. 1995. A new pressure probe method to determine the average volumetric elastic modulus of cells in plant tissues. Plant Physiology 107: 995 - 1005. 252 Nadler A, Raveh E, Yermiyahu U, Green SR. 2003. Evaluation of TDR use to monitor water content in stems of lemon trees and soils and their responses to water stress. Soil Science Society of America Journal 67: 437 - 448. Nadler A, Raveh E, Yermiyahu U, Green SR. 2006. Stress-induced water content variations in mango stem by time domain reflectometry. Soil Science Society of America Journal 70: 510 - 520. Nadler A, Raveh E, Yermiyahu U, Lado M, Nasser A, Barak M, Green S. 2008. Detecting Water Stress in Trees Using Stem Electrical Conductivity Measurements. Soil Science Society of America Journal 72: 1014 - 1024. Nadler A, Tyree MT. 2008. Substituting Stem’s Water Content by Electrical Conductivity for Monitoring Water Status Changes. Soil Science Society of America Journal 72: 1006 - 1013. Naor A, Gal Y, Peres M. 2006. The inherent variability of water stress indicators in apple, nectarine and pear orchards, and the validity of a leaf-selection procedure for water potential measurements. Irrigation Science 24: 129 - 135. Naor A, Naschitz S, Peres M, Gal Y. 2008. Responses of apple fruit size to tree water status and crop load. Tree Physiology 28: 1255 - 1261. Nardini A, Pedá G, Salleo S. 2012. Alternative methods for scaling leaf hydraulic conductance offer new insights into the structure - function relationships of sun and shade leaves. Functional Plant Biology 39: 394 - 401 Neill S. 2007. Interactions between abscisic acid, hydrogen peroxide and nitric oxide mediate survival responses during water stress. New Phytologist 175: 4 - 6. Netzer Y, Yao C, Shenker M, Bravdo B, Schwartz A. 2009. The use of crop coefficients for skilled irrigation of table grapes trained to a large trellis system. Irrigation Science 27: 109 - 120. Neumann P. 2008. Coping mechanisms for crop plants in drought-prone environments. Annals of Botany 101: 901 - 907. Nonami H, Schulze ED. 1989. Cell water potential, osmotic potential, and turgor in the epidermis and mesophyll of transpiring leaves. Planta 177, 35 - 46. 253 Nortes PA, Gonzalez-Real MM, Egea G, Baille A. 2009. Seasonal effects of deficit irrigation on leaf photosynthetic traits of fruiting and non-fruiting shoots in almond trees. Tree Physiology 29: 375 - 388. Oren R, Sperry JS, Katul GG, Pataki DE, Ewers BE, Phillips N, Schafer KVR. 1999. Survey and synthesis of intraand interspecific variation in stomatal sensitivity to vapour pressure deficit. Plant, Cell and Environment 22: 1515 - 1526. Orgaz F, Fereres E. 1997. Riego. In: Barranco, D., Fernandez-Escobar, R., Rallo, L. (Eds.), El cultivo del olivo. Mundi Prensa, Madrid, pp. 251 - 272. Orgaz F, Testi L, Villalobos FJ, Fereres E. 2006. Water requirements of olive orchards II: determination of crop coefficients for irrigation scheduling. Irrigation Science 24: 77 - 84. Ortuño MF, Conejero W, Moreno F, Moriana A, Intrigliolo DS, Biel C, Mellisho CD, Pérez-Pastor A, Domingo R, Ruiz-Sánchez MC. 2010. Could trunk diameter sensors be used in woody crops for irrigation scheduling? A review of current knowledge and future perspectives. Agricultural Water Management 97: 1 - 11. Ortuño MF, García-Orellana Y, Conejero W, Ruiz-Sánchez MC, Alarcón JJ, Torrecillas A. 2006. Stem and leaf water potentials, gas exchange, sap flow, and trunk diameter fluctuations for detecting water stress in lemon trees. Trees 20: 1 - 8. Ounapuu E, Sellin A. 2013. Daily dynamics of leaf and soil-to-branch hydraulic conductance in silver birch (Betula pendula) measured in situ. Plant Physiology & Biochemistry 68: 104 - 110. Oyarce P, Gurovich L. 2011. Evidence for the transmission of information through electric potentials in injured avocado trees. Journal of Plant Physiology 168: 103 - 108. Palomo M, Moreno F, Fernández J, Dıaz-Espejo A, Girón I. 2002. Determining water consumption in olive orchards using the water balance approach. Agricultural Water Management 55: 15 - 35. Pantin F, Monnet F, Jannaud D, Costa JM, Renaud J, Muller B, Simonneau T, Genty B. 2012. The dual effect of abscisic acid on stomata. The New Phytologist 197: 65 - 72. 260 2009. A non-invasive probe for online-monitoring of turgor pressure changes under field conditions. Plant Biology 11: 701 - 712. Wistuba N, Reich R, Wagner H-J, Zhu JJ, Schneider H, Bentrup F-W, Haase A, Zimmermann U. 2000. Xylem Flow and its Driving Forces in a Tropical Liana: Concomitant Flow-Sensitive NMR Imaging and Pressure Probe Measurements. Plant Biology 2: 579 - 582. Woodward FI. 1998. Do plants really need stomata? Journal of Experimental Botany 49: 471 - 480. Xu S-M, Liu L-X, Woo KC, Wang D-L. 2007. Changes in photosynthesis, xanthophyll cycle, and sugar accumulation in two North Australia tropical species differing in leaf angles. Photosynthetica 45: 348 - 354. Yao C, Moreshet S, Aloni B. 2001. Water relations and hydraulic control of stomatal behaviour in bell pepper plant in partial soil drying. Plant, Cell & Environment 24: 227 - 235. Zeder MA. 2008. Domestication and early agriculture in the Mediterranean Basin: origins, diffusion, and impact. Proceedings of the National Academy of Sciences of the United States of America 105: 11597 - 11604. Zimmermann D, Reuss R, Westhoff M, Gessner P, Bauer W, Bamberg E, Bentrup F-W, Zimmermann U. 2008. A novel, non-invasive, online-monitoring, versatile and easy plant-based probe for measuring leaf water status. Journal of Experimental Botany 59: 3157 - 3167. Zimmermann D, Westhoff M, Zimmermann G, Gessner P, Gessner A, Wegner LH, Rokitta M, Ache P, Schneider H, Vásquez JA, et al. 2007. Foliar water supply of tall trees: evidence for mucilage-facilitated moisture uptake from the atmosphere and the impact on pressure bomb measurements. Protoplasma 232: 11 - 34. Zimmermann U, Rüger S, Shapira O, Westhoff M, Wegner LH, Reuss R, Gessner P, Zimmermann G, Israeli Y, Zhou A, et al. 2009. Effects of environmental parameters and irrigation on the turgor pressure of banana plants measured using the non-invasive, online monitoring leaf patch clamp pressure probe. Plant Biology 12: 424 - 436. 261 Zimmermann MH. 1983. Xylem Structure and the Ascent of Sap. Springer, Berlin. Zimmermann U, Schneider H, Wegner LH, Haase A. 2004. Water ascent in tall trees: does evolution of land plants rely on a highly metastable state? New Phytologist 162: 575 - 615. Zufferey V, Cochard H, Ameglio T, Spring J-L, Viret O. 2011. Diurnal cycles of embolism formation and repair in petioles of grapevine (Vitis vinifera cv. Chasselas). Journal of Experimental Botany 62: 3885 - 3894. Zweifel R, Item H, Häsler R. 2000. Stem radius changes and their relation to stored water in stems of young Norway spruce trees. Trees 15: 50 - 57. Zweifel R, Item H, Häsler R. 2001. Link between diurnal stem radius changes and tree water relations. Tree Physiology 21: 869 - 877.