Branch sap flow in a mature olive tree: Dynamics and relation to architectural traits

This study examined how sap flow in shoots and branches of a mature olive tree (Olea europaea ‘Coratina’) related to morphological traits which are linked to tree hydraulic architecture. Spatial and temporal variations of sap flow within the tree canopy were investigated; they were analyzed with respect to the ratio between sapwood area and leaf area (SA/LA), and in relation to evaporation demand. The work was carried out on distinct sets of sap flow data, collected using different thermal methods (Heat Field Deformation HFD and Stem Heat Balance with external heating – SHB). Sap flow density (q) appeared to be governed largely by the ratio SA/LA, irrespective of the strength and heterogeneity of the evaporation demand that affected the tree water loss. The relationship between q and SA/LA suggested a higher efficiency of the water supply pathway when SA/LA was lower, i.e. in smaller (and more distal) branches. INTRODUCTION In vascular plants the supply of water to sustain transpiration (and associated carbon uptake) depends on soil water availability, on the extent of transpiring and absorbing surfaces and on the transport characteristics of the hydraulic pathway between roots and leaves. Few studies have attempted to describe hydraulic architecture and properties in species of Mediterranean ecosystems. Thompson et al. (1983) and Salleo et al. (1985) studied hydraulic parameters in stems and twigs of very young olive trees. Tognetti et al. (2005) determined seasonal courses of whole-plant hydraulic conductance in an olive grove subjected to deficit irrigation. We studied an individual tree in an old olive grove (southern Italy) to examine how sap flow in different branch ranks related to tree architectural traits. Multiple measurements of sap flow at various positions within the crown were related to the ratio between sapwood area (i.e. the cross-sectional area of conducting tissue, SA) and the transpiring surface (leaf area, LA). The ratio SA/LA (the Huber value, Zimmermann, 1983) is a morphological index of potential capacity for water transport relative to potential transpirational demand in the crown. SA/LA is an attribute that governs hydraulic architecture; in tropical forest trees it appeared to be a reliable proxy for higher order components of tree hydraulic architecture, namely leafand sapwood-related hydraulic conductivities (Meinzer et al., 2008). The relationship between the sap flow density (q, g cm h) and the ratio SA/LA was determined and analyzed. First we studied the relationship between the average diurnal sap flow (qavg, calculated from 5 am to 7 pm) and the SA/LA ratio, in the different structures (branches and shoots). Subsequently we examined the relationship between the maximum daily values of sap flow (qmax) in the distinct structures and SA/LA. Finally we studied the relationship between sap flow density and SA/LA when branches and shoots a [email protected] 315 Proc. VII th IW on Sap Flow Eds.: E. Fernández and A. Diaz-Espejo Acta Hort. 846, ISHS 2009 undergo heterogeneous atmospheric evaporation demand. MATERIALS AND METHODS The Experimental Site and the Sample Tree Data were collected in 2002, from July 22 to August 3, during the Intensive Observation Period 1 of the EU project WATERUSE (http://www.isa.utl.pt/wateruse). The field campaign was carried out in an olive grove (Olea europaea ‘Coratina’), located at the farm "Torre di Bocca", near Andria, on the "Murge" upland of Puglia region, in southern Italy (41°13' N, 16°09' E, altitude 170 m). The area is characterised as semi-arid, with average annual rainfall of 530 mm, distributed from September to April. The orchard (1 ha) was drip-irrigated. Plants (approximately 80 years old) were trained with the vase system and planted at a distance of 9 m x 9 m, with a ground cover of 23%. Typically, the trees were about 5 m high and mean crown radius was 2.4 m. The stand leaf area index (LAIstand) was low (0.96 m m), indicating minimal mutual shading by neighbouring trees (limited, in fact, to early morning and late evening hours). The leaf area index values of individual trees (LAItree), related to their crown projected area, resulted on average 3.5 m m (in a range from 1 to 7 from the middle crown to the maximum along crown radius, Cermak et al., 2007a). This resulted in usual leaf selfshading within tree crowns. One representative sample tree was selected in the middle of the olive grove; from the biometric viewpoint it approached the mean tree of the stand (diameter at 1 m height = 36 cm). The tree had a single trunk with 3 main branches directed to East, West and North. Measurements of Sap Flow and Environmental Data Distinct sets of sap flow data were collected using different thermal methods. Measurements of sap flow density (q, g cm h) in branches were performed by the Heat Field Deformation technique (HFD, Nadezhdina et al., 1998; Cermak et al., 2004). In shoots sap flow rate (g h) was measured by the Stem Heat Balance sensors (SHB), developed by Sakuratani (1981) and produced by Dynamax Inc. q in shoots was calculated according to the sapwood cross-sectional area (SA, cm) at sensor level. Signals from the sap flow probes were scanned every 10 s and their 10-minute averages were stored in data loggers (CR10, Campbell Sci., UK; Delta-T Ltd., UK; MIDI-12, EMS, CZ). Data of two days (= 15% of total) were chosen for detailed analysis. Thirteen sensors were placed in various crown positions, at different heights and orientations, and in distinct branch ranks. On the three main branches (diameter of 12, 11.5, and 10 cm) six HFD multi-point sensors were placed; measurements in smaller higher branches were performed by means of four single-point HFD sensors. Stem Heat Balance sensors (models SGA13 and SGB16) were placed at the base of three shoots. The main traits of the branches studied are illustrated in Table 1, and the scheme of sensors’ installation is shown in Figure 1. Net radiation was measured by a radiometer (CNR1, Kipp & Zonen, NL) placed on a micrometeorological tower located in the olive grove, some 20 m away from the sample tree. Volumetric soil water content was determined once a day through a TDR-system (Tektronix 1502C). The profile of soil water content was measured by means of four probes placed horizontally at four depths, down to 1 m. Two sets of four probes were arranged near the sample tree, at distances of 2 and 4.5 meters from the trunk. Leaf Area and Sapwood Area Estimates The leaf area of the gauged shoots was estimated from the number of leaves above the sensor and mean area per leaf. Branch leaf area was estimated from allometric relationships (determined at the same site) with branch xylem cross-sectional area as an independent and easily measurable variable (Cermak et al., 2007a). In the main branches the HFD probes were 40 mm long and contained 6