# Source paper archive - Source title: Phloem conductance from leaf to fruits in avocado during development - Author(s): Agullo-Samper, M., Ferrer-Blanco, C., Gonzalez, C. E., Gennis, S., Jensen, K. H., Losada, J. M. - Source landing URL: https://www.biorxiv.org/content/10.64898/2026.10.31.762750v1 - PDF URL: https://www.biorxiv.org/content/10.64898/2026.10.31.762750v1.full.pdf - PDF SHA-256: 4c85f7e22443f04a557a3e8da002a601f7388f1509d88113793d0623328494c4 - Converter: pymupdf4llm 0.0.27 - Limits: Automated PDF text extraction; reading order, equations, tables and formatting may be incomplete or incorrect. Images are omitted, and no OCR was performed. Markup and links are neutralized for safe downloading. Consult the original PDF for the authoritative version. --- [bioRxiv preprint doi: https://doi.org/10.64898/2026.10.31.762750; this version posted November 2, 2026. The copyright holder for this](https://doi.org/10.64898/2026.10.31.762750) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. # 1 Title: 2 Phloem conductance from leaf to fruits in avocado during development # 3 Short title: 4 Phloem conductance in avocado # 5 Authors: 6 María Agulló-Samper [1], Cristina Ferrer-Blanco [1], Carlos E. González [1], Sabrina Gennis [3], 7 Kaare H. Jensen [3], Juan M. Losada [1* ] 8 1 Institute for Mediterranean and Subtropical Horticulture ‘La Mayora’ (IHSM LaMayora 9 CSIC-UMA). Avda. Dr. Wienberg s/n. Algarrobo-Costa, 29750, Málaga, Spain. 10 2 Department of Physics, Technical University of Denmark, DK-2800 Kgs. Lyngby, 11 Denmark 12 - **Author for correspondence:** 13 _Juan M. Losada_ 14 _Telephone: +34 952548990_ 15 _E-mail:_ _[juan.losada@csic.es](mailto:juan.losada@csic.es)_ . 16 17 **Author contributions: MAS:** Formal analysis, methodology phloem anatomy, 18 Visualization, Writing – review and editing. **CFB:** Formal analysis, methodology of 19 NSCs visualization, writing – review and editing **CEG:** Formal analysis, Visualization 20 STE number, **SG:** Conceptualization of modeling, Formal analysis, Methodology, 21 Visualization of results, Writing – review and editing, **KHJ:** Conceptualization of 22 modeling, Formal analysis, Methodology, Visualization of results, Writing – review and 23 editing. **JML:** Conceptualization of the work, Formal analysis of data, Visualization of 24 figures, Writing – original draft. 1 [bioRxiv preprint doi: https://doi.org/10.64898/2026.10.31.762750; this version posted November 2, 2026. The copyright holder for this](https://doi.org/10.64898/2026.10.31.762750) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 1 **ABSTRACT** 2 Reproductive function in seed plants depends on efficiently allocating nutrients from 3 sources to developing fruits. Hydraulic transport of carbohydrates largely depends on 4 phloem architecture, but how phloem structure in fruit pedicels changes with 5 development remains unexplored for most seed plants. Using the one-year phenology of 6 avocado trees, we quantified water content, dry matter, and non-structural carbohydrates 7 (NSCs) in mature leaves, flowers, young fruits, and fully developed fruits. 8 Concomitantly, we measured phloem conduit geometry in different vein orders of leaves 9 and in flower-fruit pedicels over time, including sieve tube numbers in cross sections. 10 NSCs were below 5% in leaves and fruits but reached the highest concentrations in 11 flowers, especially fructose (20%) and glucose (35%), and the pedicels of flowers showed 12 the smallest conduits. While sieve tube elements in leaves scaled their geometry from the 13 minor veins to the petioles, in fruit pedicels, their geometry changed only from flower to 14 initial fruit set, increasing their number linearly up to fruit maturation. Our models show 15 that phloem transport toward the flowers is impaired, but total conductance increases 16 gradually due to phloemogenesis. 17 18 19 20 21 22 23 24 25 26 27 **Key words:** Phloem, sinks, sources, modelling, sieve tube, avocado, fruits. 28 2 [bioRxiv preprint doi: https://doi.org/10.64898/2026.10.31.762750; this version posted November 2, 2026. The copyright holder for this](https://doi.org/10.64898/2026.10.31.762750) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. # 1 INTRODUCTION 2 The structure of the phloem in source and sink organs plays a fundamental role in resource 3 allocation in plants during development, influencing key processes such as reproductive 4 function. However, research on comparing the structure of the phloem in sources and 5 sinks remains unstudied in most vascular plants. The mass flux of photo-assimilate 6 transport and delivery are critically influenced by the geometry of the conduits 7 transporting them, which includes length, radius, sieve plate pore size and number, and 8 sieve tube number in cross-section (members of our team; Horwitz, 1954; Thompson and 9 Holbrook, 2003; Jensen _et al_ ., 2013). The scaling geometry of the individual conduits that 10 form a continuum in the phloem, or sieve tube elements (STEs), determines their 11 hydraulic efficiency (Jensen _et al_ ., 2012ab; Savage _et al_ ., 2017; Barceló-Anguiano _et al_ ., 12 2021b), but studies focused on the geometry of the phloem conduits in the different vein 13 orders have only recently emerged. So far, phloem conduit scaling in leaves has been 14 studied in a limited number of species that include different growth forms, such as lianas 15 (Losada _et al_ ., 2019), shrubs (Losada & Holbrook, 2019), and trees (Carvalho _et al_ ., 2018; 16 Barceló-Anguiano _et al_ ., 2021a,b; He _et al_, 2024). All these works converge on the 17 scaling relationship between vein order and sieve tube element size. Yet, the number of 18 conduits, which plays a critical role in the bulk mass flow at the organ level, remains 19 unexplored in most species with reticulate venation, except for leaves with a single vein, 20 such as the needles of pines (Rosenberg _et al_ ., 2022; see also Vincent _et al_ ., 2025). The 21 lack of information regarding the number of conduits in other species limits our 22 understanding of net photo-assimilate transport from source to sink organs. 23 Developmentally, the mature leaf reached the maximum differentiation of 24 vascular bundles; thus, it can be considered static (Esau, 1969). In contrast, the continuous 25 gain of matter during fruit development putatively requires differentiation of the vascular 26 cambium into more conductive phloem, but whether the number of conduits changes in 27 fruit pedicels over time remains unexplored. At the same time, vascular flow rates have 28 been measured mainly in the xylem of fruit pedicels, revealing a direct relationship with 29 dry matter accumulation in several fruit trees, like kiwifruit (Clearwater _et al_ ., 2012; 30 Morandi _et al_ ., 2010), grape (Greenspan _et al_ ., 1994), apple (Lang, 1980), peach (Morandi 31 _et al_ ., 2007), pear (Morandi _et al_ ., 2014), sweet cherry (Brüggenwirth _et al_ ., 2016; 32 Winkler _et al_ ., 2016), and avocado (Kaneko _et al_ .,; 2024). Yet, these studies focus on 33 functional outcomes (water/flow rates) rather than the underlying structural constraints 3 [bioRxiv preprint doi: https://doi.org/10.64898/2026.10.31.762750; this version posted November 2, 2026. The copyright holder for this](https://doi.org/10.64898/2026.10.31.762750) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 1 of the phloem conduits (size and number) that ultimately determine the long-term 2 hydraulic capacity for resource allocation. 3 While water drives carbohydrates towards fruits, they can either be stored, as in 4 grapes, or transformed into lipids, as in avocados. Carbohydrates, synthesized in the foliar 5 cells of the mesophyll, are loaded into the phloem conduits of the minor veins, increasing 6 their osmotic pressure, which allows the passive entrance of water from the xylem, and 7 that initiates a flow through the central veins, the petiole, the branches, eventually 8 unloading in the growing organs, where sugar unloading occurs, dropping the osmotic 9 pressure and thus allowing for directional transport (Münch, 1930). While daily changes 10 in carbohydrate transport are well described (Brauner _et al_ ., 2018), little is known about 11 seasonal changes that govern fructification. The rate of vascular flow changes seasonally 12 and is constrained by the architectural construction of the phloem conduits (size and 13 number) over time (Vincent _et al_ ., 2025). However, only a few studies have examined the 14 structure and function of the phloem in developing fruits (Savage _et al_ ., 2015), and no 15 quantification of the individual phloem elements in developing fruits has been reported 16 to date. This is because distinguishing the sieve tube elements of the phloem requires a 17 combination of microscopy and immunocytochemistry (Torode _et al_ ., 2018), which are 18 typically costly and time-consuming. Comparing the morphology and number of sieve 19 tube elements between petioles (sources) and pedicels (sinks) is essential for 20 understanding the critical developmental stages of plant reproduction, with implications 21 for ecology and agriculture. This research aims to shed light on how the vascular system 22 of avocado pedicels changes over time during fruit growth, compared with that of 23 petioles. 24 Avocado is increasingly becoming a staple food worldwide, and its cultivation has 25 been rising by more than 100% in the last decade, with Mexico as the top producer. In 26 Continental Europe, where the market of avocado grew 16% in 2025, Spain is the major 27 producer, centered in the Mediterranean coast of the south, but expanding to other coastal 28 areas (Rodríguez Pleguezuelo _et al_ ., 2018). It is also one of the highest latitudes for 29 avocado cultivation, which influences their productivity rates, with one fruit per year 30 compared with the two- or three-sometimes fruit production in tropical areas of the world 31 (López-Hoyos _et al_ ., 2024). Due to its growing economic and social impact, numerous 32 studies have focused on breeding and agronomic improvement of avocado fruits, but 33 challenges of climate change require a physiological approach. Recent physiological 4 [bioRxiv preprint doi: https://doi.org/10.64898/2026.10.31.762750; this version posted November 2, 2026. The copyright holder for this](https://doi.org/10.64898/2026.10.31.762750) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 1 studies on avocado focused on the hydraulics of the xylem under drought (Cardoso _et al_ ., 2 2020) or on the evaluation of vascular flows toward the fruits (Kaneko _et al_ ., 2022). Yet, 3 there are no reports on the anatomical structure of avocado pedicels, even though it is 4 crucial to understand the hydraulics of sugar transport to growing fruits. We used the most 5 widely consumed variety worldwide, ´Hass´, to evaluate non-structural carbohydrates in 6 leaves and fruits over time, and to compare phloem structure in leaf petioles and fruit 7 pedicels. Our results reveal that phloem differentiation in fruits plays a critical part in the 8 hydraulic transport of nutrients toward developing fruits. # 9 RESULTS 10 **Phenology of avocado cv. `Hass´ vegetative and reproductive organs** 11 The developmental cycle of avocado trees in the subtropical south of Spain takes about 12 months 12 ( **Figure 1** ; also described in Alcaraz _et al_ ., 2013). Following the mild winter, almost all leaves 13 drop off, concomitant with the emergence of inflorescences around March-April ( **Figure 1A** ). 14 During the pollination period, inflorescences containing tens of tiny yellow flowers at the tips of 15 the branches (hundreds per tree) coexist with a few developing leaves of the current year ( **Figures** 16 **1B-C** ). Following pollination, a small proportion of the flowers set fruit (Cameron _et al_ . 1952), 17 and reached the maximum percentage of water (80%) and around 18% of dry matter (Fig S1). 18 Concomitantly, leaves of the first flush extend over a period of approximately 30 days (until the 19 end of May), when they reach a complete expansion ( **Figures 1D-F** ). Fruits took four months 20 (October, month 10) to reach complete expansion, and two months later (month 12), they reached 21 full maturity and displayed a reduction in water content to 70% and an increase in dry matter to 22 28% ( **Supplemental Figure S1** ). Water percentage was minimal in flowers (10%), contrasting 23 with the 70% in leaves and developing fruits, which reached a maximum of water during the 24 initial stages of fruit set ( **Figure1G** ). 25 The length-to-width ratio of reproductive organs increased from below 1.5 in flowers to above 26 this value during the initial fruit set ( **Figure 2A** ). Still, it remained between 1.5 and 2.0 from the 27 fruitlets to the mature fruits, without significant differences. Similarly, the diameter of the flower 28 pedicels was 2mm ( **Figure 2B** ), and increased significantly at the initial fruit set in July, reaching 29 a maximum of 5mm. Still, this value remained unchanged for months, until the fruits reached 30 maximum expansion in October and subsequently matured in December. 5 [bioRxiv preprint doi: https://doi.org/10.64898/2026.10.31.762750; this version posted November 2, 2026. The copyright holder for this](https://doi.org/10.64898/2026.10.31.762750) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 1 **Non-structural carbohydrates in avocado leaves and fruits during development** 2 Comparison of starch in fully developed leaves from the previous and the current year revealed a 3 similar amount of starch, around 5% ( **Figure 3A** ). Flowers reached a peak of ~9% of starch, 4 which was later reduced significantly to 2% in developing fruits, reaching a minimum of 0.5% in 5 the mature fruits. Sucrose content was below 5% in all organs and fruit developmental stages, 6 with little differences among them ( **Figure 3B** ). The content of fructose and glucose followed 7 similar patterns, with 5% or lower in the leaves, a maximum amount in the flowers (20% and 35% 8 w/w respectively), which decreased to the developing fruits, reaching a minimum of ~1% in the 9 mature fruits ( **Figures 3C, D** ). 10 **Phloem anatomy in the petioles and pedicels of developing avocado fruits** 11 _**Length and width**_ 12 Evaluations of the sieve tube elements in the different vein orders of the avocado leaves ( **Figure** 13 **4A** ) revealed a scaling relationship that was in line with the vein size (i.e. vein order). Thus, the 14 minor veins showed the shortest and thinnest sieve tube elements, which increased in length to 15 the veins or higher orders ( **Figures 4B-C** ). The STE length in the petiole was smaller than that in 16 the primary vein, and similar to the secondary order veins. Similarly, the width of the sieve tube 17 elements increased gradually from minor to major veins, and the sieve tubes of the petiole and 18 the major veins were similar in diameter. 19 Measurements of the phloem in the pedicels of reproductive organs revealed that the smallest 20 sieve tube element width corresponded with the pedicels of flowers, increasing slightly to 21 developing fruits (July or month 7 of development), maintaining the same values until maturity 22 in December ( **Figure 5A** ). The length of the sieve tube elements was also minimum at the flower 23 stage, and increased some microns significantly from flower to fruit development and maturity 24 ( **Figure 5B** ). 25 _**Sieve plate pore size and number**_ 26 To understand whether the sieve plate connections followed this pattern ( **Figure 6A** ), the average 27 radius of the pores connecting the sieve plate was minimum in the flowers (around 0.2µm), 28 doubling this value in developing fruits (0.4µm), but maintaining this size up to the mature fruits 29 ( **Figure 6B** ). Strikingly, the average number of pores per plate was similar in all the studied 30 developmental stages, without significant differences between flowers, developing and mature 31 fruits, despite the size differences and the prolonged temporal interval of 10 months between both 32 stages ( **Figure 6C** ). Maintenance of sieve tube element features during fruit development led to 33 quantifying their number. 6 [bioRxiv preprint doi: https://doi.org/10.64898/2026.10.31.762750; this version posted November 2, 2026. The copyright holder for this](https://doi.org/10.64898/2026.10.31.762750) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 1 _**Sieve tube element number**_ 2 Cross sections of the petioles and the pedicels revealed that the size of the axial areas of the 3 phloem correlated directly with the number of sieve tube elements in both peduncles of sources 4 and sinks (including flowers, developing fruits, and mature fruits) ( **Figure 7A** ). The size of axial 5 areas increased linearly with development, reaching a maximum at the mature stage of the fruits. 6 The number of tubes followed a similar pattern. It changed from ±6,000 tubes in the pedicel of 7 flowers to nearly ±20,000 tubes in mature fruits ( **Figure 7B** ). Strikingly, the density of sieve tube 8 elements (number per unit area) was maintained constant during all developmental stages, varying 9 between 9,000 and 10,000 per mm [2] . 10 **Conductivity and conductance in the source and sink organs of avocado** 11 Flow of sugar through the sieve tube elements of the phloem can be described by the pressure 12 driven flow hypothesis (Horwitz, 1954; Münch 1930; Jensen _et al_ ., 2013, Knoblauch _et al_ ., 2016), 13 using as a model the Hagen-Poiseuille equation (Thompson & Holbrook, 2003) 14 𝑄= (1) 𝑅 [𝛥𝑝=] 𝜂𝐿 𝑡𝑜𝑡 [𝛥𝑝][. ] 𝑄= 1 𝑅 [𝛥𝑝=] 𝑘𝜋𝑟 [2] 15 where _Q_ is the volumetric flow rate, 𝛥 _p_ the osmotic pressure difference between source and sink, 16 _R_ the hydraulic resistance, _k_ the hydraulic conductivity, 𝜂 the viscosity of the fluid, _r_ the radius 17 of the total conductive area, and _L_ tot the length of the conducting tube. Another measure is the 18 flux, which is invariant of the total cross-sectional area as it is calculated to 𝑞= 𝑄/(𝜋𝑟 [2] ). 19 Hydraulic conductivity is a widely used measure of fluid flow in the phloem and can be calculated 20 from anatomical data (Thompson & Holbrook 2003b, Mullendore 2010, Jensen 2012b). 21 Following equation (1), the conductivity can be defined as 𝑘= 𝜂𝐿 𝜋𝑎 [2] 1 22 𝑘= (2) 𝜋𝑎 [2] 𝑅 𝑆𝐸 ~~[,]~~ 23 where _R_ SE is the hydraulic resistance of one sieve element, _L_ the length of the sieve element, and 24 _a_ its radius. The hydraulic resistance has been mathematically described before as the sum of the 25 lumen and plate resistance, making use of the Hagen-Poiseuille equation (Jensen 2014). 8𝜂𝐿 3𝜂 1 8𝑡 1 26 𝑅 𝑆𝐸 = 𝜋𝑎 [4] [+] 𝑁𝑏 [3] ~~[ (]~~ 1+3( [𝜎] ~~[)]~~ 2 [+] 3𝜋𝑏 1+6( [𝜎] ~~[)]~~ 2 +3( [𝜎] ~~[)]~~ 4 [)] (3) 8𝑡 2 [+] 1 𝑅 𝑆𝐸 = 8𝜂𝐿 𝜋𝑎 [4] [+] 3𝜂 𝑁𝑏 [3] ~~[ (]~~ 3𝜂 1+3( [𝜎] 𝑏 ~~[)]~~ 1 3𝜋𝑏 [𝜎] 𝑏 ~~[)]~~ 1+6( [𝜎] 𝑏 ~~[)]~~ 2 +3( [𝜎] 𝑏 27 where _t_ is the thickness of the sieve plate, and _b_ is the average pore radius with the standard 28 deviation 𝜎. While the conductivity is specific to one sieve element and can be used to compare 29 different plants, the total conductance _C_, defined as _C_ = 1 / _R,_ allows us to consider the number of 30 sieve elements per cross-section. Similar to Jensen _et al_ . (2012) defining the resistance of one 7 [bioRxiv preprint doi: https://doi.org/10.64898/2026.10.31.762750; this version posted November 2, 2026. The copyright holder for this](https://doi.org/10.64898/2026.10.31.762750) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 1 sieve element, we can apply the electrical circuit theory to add multiple sieve elements in one 2 cross-section by adding their resistance in parallel (1/ _R_ = 1/ _R_ 1 +1/ _R_ 2 ), and various elements in the 3 longitudinal direction will be added as a serial connection ( _R_ = _R_ 1 + _R_ 2 ). Under the assumption 4 that the phloem geometry is the same along the whole translocation length _L_ tot, the total resistance 5 ( _R_ _3D_ ) is 6 𝑅 3𝐷 = (4) 𝐿 𝑛 ℎ [𝑅] [𝑆𝐸] [,] 𝑅 3𝐷 = 𝐿 𝑡𝑜𝑡 𝐿 1 7 where _n_ _h_ is the number of sieve elements per cross-section. Similarly, the conductance is 8 𝐶= 𝑛 ℎ (5) 𝐿 𝑡𝑜𝑡 𝑅 𝑆𝐸 [.] 𝐶= 𝐿 𝐿 𝑡𝑜𝑡 9 Using eqns. (2) and (3), we calculated the hydraulic conductivity in the avocado leaf as well as in 10 the flower and the fruit ( **Figure 8** ), where two developmental stages were chosen, month 7 (initial 11 fruit set) and month 12 (mature fruits). Within the veins of an avocado leaf, the conductivity of 12 the sieve elements changes by about 2/3 from 0.3 to 0.45 µm [2] . The conductivity in the flower is 13 much lower than in the leaf, with approximately 0.1 µm [2] ( **Figure 8** ). During the development of 14 the fruit, the conductivity increases by 150 % from flower to young fruit at the month 7 of the 15 year. Then it increases further by an additional 60 % to the mature fruit at the month 12, where a 16 hydraulic conductivity of about 0.4 µm [2] is reached. 17 Next to the conductivity, the total phloem conductance ( _C_ _tot_ ) was determined at the different 18 developmental stages of the fruit pedicel ( **Figure 9A** ) using eqn. (5) together with the 19 experimental number of conduits per cross-section (n h, see **Figure 7** ) and an initial estimation of 20 1 m for the distance between source and sink (L tot ). The conductance of the phloem increased 21 from flower to fruit and from developing to mature fruits. This growth is mainly explained by the 22 increase in sieve element number, which scales proportionally with the conductance (eqn. (5)). 23 To verify the calculated values for biological relevance, we further calculated the pressure drop 24 that would result from the conductance ( **Figure 9B** ). The pressure gradient to drive the viscous 25 flow in a plant is of the order of 0.1-1 MPa (Jensen _et al_ ., 2013). We therefore expect a pressure 26 gradient within the same range. The relationship between conductance, _C_, and pressure drop, _∆p_, 27 is given by eqn. (1) (keeping in mind that _C_ = 1/R). 28 To estimate the sugar import flow rate _Q,_ we rely on measurements of the fruit growth rate _G_ 29 ( **Supplemental Figure S1** ) 30 𝐺= 𝜌𝑄, (6) 31 where 𝜌 is the density of the phloem sap. Here, we assume a typical sucrose concentration of 18% 32 wt (Jensen _et al_ ., 2013). 8 [bioRxiv preprint doi: https://doi.org/10.64898/2026.10.31.762750; this version posted November 2, 2026. The copyright holder for this](https://doi.org/10.64898/2026.10.31.762750) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 1 From Eqns. (4) and (5), the pressure gradient can be evaluated as a function of the source-to-sink 2 distance 𝐿 𝑡𝑜𝑡. . Here, we focus on two examples: 𝐿 𝑡𝑜𝑡 = 0.1m and 1m. The inferred pressure 3 gradients generally fall within or below the typically observed range (0.1-1 MPa, see Jensen _et_ 4 _al_ ., 2013). While values below 0.1 MPa are not usually observed, they are physically achievable 5 by the plant, unlike pressure gradients above 1 MPa. Variations to the pressure gradients 6 calculated here, could be explained by our general assumption of a uniform system with no change 7 in phloem geometry across the transport area between source and sink. In general, the pressure 8 drop required to achieve the observed growth decreases with increasing fruit maturity, likely due 9 to an increase in the number of sieve tubes per cross-section and, thereby, conductance ( **Figure** 10 **9B** ). 11 In contrast, we calculated the resulting flow rate at a fixed pressure drop of 0.1 MPa ( **Figure 9C** ). 12 While the results show a variation between the experimentally estimated flow rates from eqn. (6) 13 to the calculated ones, any differences might be explained by our assumptions of a uniform 14 system; however, both this result and the estimated pressure drop show that a higher flow is 15 theoretically possible. # 16 Discussion 17 This empirical study evaluates the role of phloem conduit number from source organs (leaf 18 petioles) to the pedicels of developing fruits over time, addressing a knowledge gap regarding the 19 dynamic nature of phloem architecture in developing sinks. Our findings challenge the idea of 20 continuous scaling of conduit size in fruit pedicels, instead highlighting a bimodal strategy: a 21 rapid increase in individual conduit efficiency during initial fruit set, followed by a sustained rise 22 in total conductive area through phloemogenesis. 23 **Initial bottleneck during avocado fruit development: adaptation by size** 24 Our work quantified, for the first time in a fruit tree, changes in phloem structure (sieve tube 25 element geometry, sieve plate morphology, and conduit number) in avocado pedicels from flower 26 to mature fruit over time (months). These evaluations revealed a critical initial limitation. In 27 avocado flower stalks, the measured sieve tube elements showed a minimum size (approximately 28 0.1µm radius) and the smallest pore radii, resulting in a high hydraulic resistance (Jensen _et al_ ., 29 2012; Savage _et al_ ., 2017). This low conductivity coincides with the highest nonstructural 30 carbohydrate concentration (20% of fructose and 35% glucose). Accumulation of sugars within 31 flowers indicates a transient impairment in transport, suggesting hydraulic translocation of 32 photoassimilates through the phloem to the flowers is initially limited (Wen _et al_ ., 2022; Ren _et_ 33 _al_ ., 2023). By maximizing solute concentration, the flowers maintain a steep osmotic pressure 34 gradient across the sieve plates, thus driving the viscous flow despite the structural limitation, as 9 [bioRxiv preprint doi: https://doi.org/10.64898/2026.10.31.762750; this version posted November 2, 2026. The copyright holder for this](https://doi.org/10.64898/2026.10.31.762750) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 1 predicted by the Münch model (Münch, 1930; Liesche, 2022). Furthermore, a handful of recent 2 studies quantifying phloem transport to floral tissues have reported low conductivity, as in 3 _Paeonia_ flowers (Yang _et al_ ., 2023). The anatomy of the phloem is an essential aspect for 4 understanding both the mechanical properties of the reproductive organs and their greater 5 vulnerability to extremes in the environment compared with leaves (Alves Aun _et al_ ., 2024). In 6 avocado, a significant carbon budget is implied during the flowering period, given the thousands 7 of flowers per tree in the spring, but most flowers will eventually drop off, given the low fruit set 8 in this species (Alcaraz _et al_ ., 2021). The high resistance i n the floral phloem may serve as a 9 protective mechanism that modulates the initial carbon allocation to reproductive tissues 10 (Turgeon, 2010), leaving a high osmotic pressure within flowers that drives water import during 11 the initial fruit elongation stages. 12 The 150% increase in sieve tube element conductivity observed from flower to initial fruit set 13 (young fruits at month 7) is a critical developmental adjustment. This is driven by the doubling 14 of the sieve plate pore radius and the tube itself. Given the flow´s proportionality to the fourth 15 power of the radius (Hagen-Poiseuille model), this slight increase represents a massive reduction 16 in resistance, effectively overcoming the initial hydraulic bottleneck (Thompson, 2006). This 17 rapid adjustment establishes an efficient minimum conduit size, allowing the pedicel to meet the 18 rising demand for photosynthates as the fruit begins its rapid growth phase (Falchi _et al_ ., 2020). 19 Unloading mechanisms in fruits are poorly understood, but previous works in fruit crops such as 20 grapes (Zhang _et al_ ., 2006) revealed a developmental shift from passive to active phloem loading 21 (Ma _et al_ ., 2019). While these unloading mechanisms are mainly related to physical gates of 22 phloem cells in fruits (i.e. plasmodesmata or active transporters such as sucrose transporters), the 23 role of phloem structure and its consequent hydraulics has been mostly overlooked. Our results 24 point to a developmental hydraulic shift from a highly resistant phloem during flowering to a less 25 resistant one during the initial stages of fruit development, depicted by an increase in the 26 geometrical parameters of sieve tubes in the pedicel of fruits that achieved a minimum size. 27 **Sustained growth of avocados during fruit elongation: adaptation by number** 28 Following the fruit set, there was a maintained growth of fruits until month 10, but the pedicel 29 showed a constant diameter. Dry matter accumulation over time, coincident with the continuous 30 drop of carbohydrates in the developing fruits, demonstrates that carbohydrate transport is 31 consistent with an enhanced phloem conductivity (Thompson, 2006). This is in line with previous 32 evaluations of starch accumulation in the stems of avocado, which showed a similar trend during 33 development (Whiley _et al_ ., 1996; Liu _et al_ ., 1999). While water transported through the xylem 34 is required for cell elongation during fruit growth (Marcelis and Heuvelink 2007; Zamski 2017), 35 and the import/export flow rates of water have been previously measured in avocado (Kaneko _et_ 10 [bioRxiv preprint doi: https://doi.org/10.64898/2026.10.31.762750; this version posted November 2, 2026. The copyright holder for this](https://doi.org/10.64898/2026.10.31.762750) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 1 _al_ ., 2024), the structural capacity of the phloem to support the required carbohydrate transport 2 rates has been traditionally ignored. 3 Our findings show that after the initial geometric change (month 7), the length, width, and sieve 4 plate characteristics remain statistically constant until month 12. In contrast, total phloem 5 conductance ( _C_ ) increases monotonically throughout fruit development, closely mirroring the 6 continuous increase in the number of sieve tubes (from 6000 to 20000 per fruit pedicel). The 7 maintenance of a constant sieve tube density (approximately 8000 to 10000 conduits per mm2) 8 strongly suggests the plant´s strategy is to increase the total cross-sectional area of the phloem 9 through cambial activity in the pedicel, a process termed phloemogenesis. Previous reports with 10 giant pumpkins strongly suggested that fruit size was strongly related with the capacity of phloem 11 transport (Savage _et al_ ., 2015), and the phloem variations with seasons has been mainly studied 12 in temperate species, given the shifts that accompany the formation of new leaves and fruits 13 during the spring (Ray and Savage, 2021). The seasonal effect has been less explored in evergreen 14 species such as avocado, but their prolonged fructification periods allow to understand the role of 15 biosynthesis of sieve tube elements over time as critical for a proper maturation and, therefore, an 16 efficient crop harvest. 17 This is a strategy of scaling-by-number rather than scaling-by-size. Interestingly, this approach is 18 different from source organs (leaves), where the STE size scales with vein order (Carvalho _et al_ ., 19 2018; Losada & Holbrook, 2019; Losada _et al_ ., 2022). This fundamental difference in vascular 20 developmental regulation ensures that total conductive capacity keeps pace with the fruit´s mass 21 accumulation rate, preventing the phloem from becoming a limiting factor. The continuous 22 structural differentiation reflects a dedicated resource investment to maintain the transport 23 capacity required by a high demand sink (Stanfield _et al_ ., 2024). The scaling-by-number strategy 24 aligns with our hydraulic models suggesting that increasing conduit number is highly effective 25 for boosting long distance transport (Liu _et al_ ., 2024). The linear rate of fruit growth thus 26 determines the capacity of phloem transport, suggesting a necessary synchrony driven by the sink. 27 The continuous structural differentiation implies that sink tissues require a degree of plasticity 28 (Jensen, 2018), in contrast to the more static vascular tissues of mature source organs. 29 **Conclusions and future directions** 30 Our data demonstrates that the hydraulic transport capacity of the phloem in avocado pedicels 31 adapts to increasing demands through a staged mechanism: initial size optimization, followed by 32 a sustained scaling-by-number (phloemogenesis). This finding provides essential anatomical data 33 necessary for refining whole plant carbon allocation models (Sevanto, 2018). 34 The question remains on the internal and external regulation of this phloemogenesis, and whether 35 this strategy could be generalized to other fruits with prolonged development. Future research 11 [bioRxiv preprint doi: https://doi.org/10.64898/2026.10.31.762750; this version posted November 2, 2026. The copyright holder for this](https://doi.org/10.64898/2026.10.31.762750) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 1 should focus on the regulatory signals that activate this sustained phloemogenesis in the pedicel, 2 investigating the role of hormonal gradients (e.g. auxin and gibberellin, which are known to 3 influence cambial activity, and turgor pressure feedback from the fruit, which has been implicated 4 in sinks like japanese plum (Grappadeli _et al_ ., 2019). Further comparative studies are needed to 5 determine if this bimodal scaling strategy is universal among economically important, highly 6 demanding fleshy fruits, necessary to understand developmental mechanisms governing 7 reproductive development. # 8 Methods 9 **Plant material, size, and relative water content of leaves, flowers, and fruits** 10 A subset of 41 trees (3-4m tall) belonging to an orchard of avocados cv`Hass´, >20 years 11 old, planted at 8 x 8m, irrigated on demand, that belongs to the germplasm collection of 12 the Institute for Mediterranean and Subtropical Horticulture ‘La Mayora’ – CSIC -UMA 13 (4°2ʹ24.32ʹʹN, 36°45ʹ32.75ʹʹE), which is located in the subtropical microenvironment of 14 Málaga province in southern Spain. 15 Three developmental stages of reproductive organs were selected to evaluate: flowers, 16 fruits after the first fruit drop -July, month 7-, fruits at the maximum size -October, month 17 10-, and mature fruits for harvest -December, month 12-. The diameter of the pedicel 18 sustaining developing fruits was measured with a caliper, along with the fruit's maximum 19 length and diameter. Data on the diameter of the flower pedicels and carpel size from 20 flowers were obtained from a previous work carried out in the same orchard (Alcaraz _et_ 21 _al_ ., 2010). 22 To calculate the relative water content (RWC), we collected three leaves per tree from a 23 total of 10 trees (n=30), 15 belonging to previous year leaves (December), and 15 24 belonging to newly expanded leaves (may), 11 groups of 10 flowers each (n=110 25 flowers), 11 fruits from month 7, and 11 fruits from month 12. They were all weighed 26 fresh (Fw), except for the small flowers, which were weighed in groups of 10 on a four 27 decimal scale. All samples were dried in the oven at 60ºC for at least 48 h and weighed 28 thereafter to get the dry weight (Dw). Relative water content per sample was calculated 29 following the protocol of Poorter _et al_ . (2009). In parallel, we measured dry matter content 30 using the non-invasive near-infrared (NIR) method (F-751, Felix Instruments) monthly 31 from the first fruit drop (month 7) to maturity (month 12). Note that the small size of the 32 flowers makes it impossible to measure their dry matter using this method ( **Figure** S1). 12 [bioRxiv preprint doi: https://doi.org/10.64898/2026.10.31.762750; this version posted November 2, 2026. The copyright holder for this](https://doi.org/10.64898/2026.10.31.762750) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 1 **Non-structural carbohydrate evaluation** 2 Fully developed leaves from the previous year (i.e. when the fruit of the current year are 3 mature or month 12) and fully expanded leaves from the current (i.e. at initial fruit set or 4 month 7), were cut from 15 adult trees (3 leaves per tree, n=45) and used along with a 5 subset of 10 flowers per each tree (n=150), two fruits per tree at the initial fruit set (n=30), 6 and two fruits per tree at the mature stage (n=30). Immediately after collection, the 7 material was incubated in an oven at 100°C for 1 hour to deactivate enzymes, and then 8 the samples were dehydrated in the same oven at 70°C for 3 days. Then, samples were 9 homogenized to a fine powder using a ball mill (TissueLyser, by Qiagen Inc., 10 Mississauga, ON, Canada). 11 For the quantification of non-structural carbohydrates (NSCs), a modified protocol of 12 Landhäusser _et al_ (2008) was followed. First, 30 mg of powder from each sample was 13 weighed, and NSCs were extracted using 80% ethanol. The enzymatic assay and 14 absorbance readings were performed in 96-well polystyrene plates (Thermo Scientific™ 15 Nunc MicroWell 96-Well Microplates) using a microplate reader (BioTek Instruments, 16 PowerWave XS2, USA). Glucose was directly quantified by an enzymatic assay in a 96 17 well plate mediated by hexokinase-glucose 6-phosphate dehydrogenase (Sigma G3293) 18 and linked to the production of NADH (absorbance 340 nm). The NADH produced by 19 hexokinase-glucose 6-phosphate dehydrogenase, after 1 hour and 40 minutes of 20 incubation, is directly proportional to the glucose present in the sample. The remaining 21 NSCs were converted by the enzymatic method into glucose, starting by starch digestion 22 using α-amylase from _Bacillus licheniformis_ (Sigma A4551) and amyloglucosidase from 23 _Aspergillus niger_ (Sigma Cat. ROAMYGLL); the aqueous phase with the hydrolysed 24 glucose was loaded into the 96-well plate along with hexokinase-glucose 6-phosphate 25 dehydrogenase, incubated for 1 hour 40 minutes, and finally, absorbance was measured 26 at 340 nm. Sucrose was first digested in a 96-well plate with an invertase enzyme from 27 baker’s yeast (Sigma I9274) to obtain glucose and fructose. After digestion, each sample 28 was loaded into a new 96-well plate containing hexokinase-glucose-6-phosphate 29 dehydrogenase, incubated, and measured as described above. Then, fructose 30 quantification was performed in a 96-well plate using a combination of phosphoglucose 31 isomerase (Sigma P538) and hexokinase-glucose-6-phosphate dehydrogenase 32 (absorbance at 340 nm). All calculations were performed on a dry-weight basis. 13 [bioRxiv preprint doi: https://doi.org/10.64898/2026.10.31.762750; this version posted November 2, 2026. The copyright holder for this](https://doi.org/10.64898/2026.10.31.762750) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 1 **Fluorescence microscopy for phloem evaluations** 2 To evaluate the size of sieve tube elements, longitudinal sections of vascular elements 3 (leaf veins, petioles, and flower and fruit pedicels) were fresh cut using a micro-scalpel 4 in 1xPBS with Leica S6D Stereo Zoom Microscope, resulting in approximately 1mm 5 sections, which were stained onto a microscope slide with 0.1% aniline blue in PO 4 K 3 6 (Currier and Strugger, 1956) to stain callose and 0.1% calcofluor white in 10mM KCl 7 (Hughes and McCully, 1975) to label cellulose. We measured a minimum of 500 tubes 8 per leaf vein, except for minor veins, which counted 266 tubes. The samples were 9 extracted from 10 leaves belonging to different trees, 10 flower pedicels (301 tubes), 10 10 fruit pedicels from month 7 (306 tubes), and 10 fruit pedicels from month 12 (400 tubes). 11 This procedure was also applied to transverse sections of the pedicel belonging to 10 12 mature fruits, which were used to count the number of vascular elements per axial area 13 of the phloem. The number of axial regions was further counted and measured 14 individually using image analysis. Calcofluor white identified the sieve tube elements of 15 the phloem in sections of avocado, and to confirm that counts were indeed tubes, we 16 immunolocalized the walls of sieve tube elements with the LM26 monoclonal antibody 17 (Torode _et al_ ., 2018). To this end, pedicels and petioles were frozen in liquid nitrogen, 18 then fixed in super chilled 100% ethanol to maintain their integrity, then included in 19 Technovit 8100 resin for several days or weeks, and blocks were made under anoxic 20 conditions at 4ºC. Sections of 4µm thickness were obtained with Leica EM UC7 21 ultramicrotome (Leica Microsystems). Immunolocalization was performed as described 22 before (Torode _et al_ ., 2018). While the total number of tubes was counted for most 23 developmental stages, the large size of the fruits implied that, after counting the number 24 of tubes in a subset of axial areas, we estimated the total number based on the average 25 number of tubes per area multiplied by the average number of the total regions counted. 26 The slides stained and immunolocalized were observed under Leica DM LB2 27 epifluorescence microscope using a 405-nm filter (aniline blue and calcofluor) and a 28 488nm filter for FITC belonging to immunolocalization. The microscope is equipped with 29 a camera and LAS V4.5 software. 30 **Scanning electron microscopy for sieve plate quantifications** 31 Small sections of petioles and pedicels from flowers and fruits at months 7 and 12 were 32 cut, immediately submerged in liquid nitrogen to prevent callose accumulation in the 14 [bioRxiv preprint doi: https://doi.org/10.64898/2026.10.31.762750; this version posted November 2, 2026. The copyright holder for this](https://doi.org/10.64898/2026.10.31.762750) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 1 sieve pores, and transferred to super-chilled 100% ethanol. The sections were incubated 2 first at -80ºC for 24h, and then gradually moved to -20ºC for 24h, 4ªC for 24h, and room 3 temperature. 1-2mm sectioned samples were cut on ethanol using a double-edged razor 4 blade. These sections were washed three times in distilled water to remove ethanol. A 5 0.1% proteinase K solution dissolved in 50mM Tris-HCl buffer, 1.5mM Ca [2+] acetate, and 6 8% Triton X-100, pH 8 (Mullendore _et al_ ., 2010) in a water bath at 60ºC for at least 14 7 days, replacing the mixture once a week. After 2 weeks, the cytoplasmic content was 8 digested, sections were incubated in ethanol to inactivate the enzyme, and then rinsed 9 three times in distilled water. These sections were incubated at 60ºC for 24h in a 0.1% α 10 amylase solution, which removes the starch accumulated on the sieve plates. We rinsed 11 the sections in ethanol, then 3 times in water, and finally removed all the water. 12 Lyophilization was done with a freeze-drier (CoolSafe 4-15L Freeze Dryers, LaboGene, 13 Allerod, Denmark) for 24h. Samples were mounted on SEM studs, covered with gold 14 palladium using a sputter coater (QUORUM Q 150R ES), and finally observed using 15 scanning electron microscopy (JEOL JSM-840). We measured at least 140 pores from 16 each pedicel of the fruits harvested in months 7 and 12, as well as from flowers. 17 **Image and data analysis** 18 All microscopic images were processed with IMAGEJ (National Institutes of Health, 19 [https://imagej.nih.gov/ij/). Specifically, we used it to measure the length and diameter of](https://imagej.nih.gov/ij/) 20 sieve tube elements, count sieve plates, and measure the area of each sieve plate/pores. 21 Data analysis was performed using R (version 4.2.1). All data followed a normal 22 distribution, as determined by a Shapiro test. ANOVA of 1 and 2 factors and Tukey’s test 23 at a _p_ <0.05 were used to test anatomical variation within pedicels, petioles, fruit size, and 24 sieve pores. # 25 FUNDING INFORMATION 26 MAS was supported by a Garantía Juvenil Ref. GARJUV_AND_21_02180, JML was funded 27 by Agencia Estatal de Investigación (grand number PID2024-162914OB-I00) European 28 Horizon SNS Project 6G-PATH Pilots and Trials through Europe, and PIDI-2024-02642 from 29 Junta de Andalucía. SG was supported by the Independent Research Fund Denmark, grant 30 number 1127-00219B. 15 [bioRxiv preprint doi: https://doi.org/10.64898/2026.10.31.762750; this version posted November 2, 2026. The copyright holder for this](https://doi.org/10.64898/2026.10.31.762750) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. # 1 Competing interests 2 Authors declare no competing interests. # 3 REFERENCES 4 **Alcaraz ML, Hormaza JI** (2021) Fruit set in avocado: pollen limitation, pollen 5 load size, and selective fruit abortion. Agronomy **11** : 1603 6 **Alcaraz ML, Thorp TG, Hormaza JI** (2013) Phenological growth stages of 7 avocado (Persea americana) according to the BBCH scale. Sci Hortic **164** : 434–439 8 **Aun MA, Farnese F, Loram-Lourenço L, de Abreu IMPG, Silva BRA,** 9 **Freitas JCE, Filho VMA, Silva FG, Franco AC, Hammond WM, Cochard H,** 10 **Menezes-Silva PE** (2024) Evidence of combined flower thermal and drought 11 vulnerabilities portends reproductive failure under hotter-drought conditions. Plant Cell 12 Environ **47** : 1971–1986 13 **Barceló-Anguiano M, Holbrook NM, Hormaza JI, Losada JM** (2021b) 14 Changes in ploidy affect vascular allometry and hydraulic function in _Mangifera indica_ 15 trees. Plant J **108** : 541–554 16 **Barceló-Anguiano M, Hormaza JI, Losada JM** (2021a) Conductivity of the 17 phloem in mango ( _Mangifera indica_ L.). Hortic Res **8** : 8 18 **Brauner K, Birami B, Brauner HA, Heyer AG** (2018) Diurnal periodicity of 19 assimilate transport shapes resource allocation and whole-plant carbon balance. Plant J 20 **94** : 776–789 21 **Brüggenwirth M, Winkler A, Knoche M** (2016) Xylem, phloem, and 22 transpiration flows in developing sweet cherry fruit. Trees **30** : 1821–1830 23 **Cameron SH, Mueller RT, Wallace A** (1952) Nutrient composition and 24 seasonal losses of avocado trees. Calif Avocado Soc Yearb **36** : 201–209 25 **Cardoso AA, Batz TA, McAdam SA** (2020) Xylem embolism resistance 26 determines leaf mortality during drought in _Persea americana_ . Plant Physiol **182** : 547– 27 554 16 [bioRxiv preprint doi: https://doi.org/10.64898/2026.10.31.762750; this version posted November 2, 2026. The copyright holder for this](https://doi.org/10.64898/2026.10.31.762750) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 1 **Carvalho M, Jensen KH** (2018) Vein scaling and hydraulic conductance in 2 angiosperm leaves. New Phytol **218** : 163–174 3 **Clearwater MJ, Luo Z, Ong SEC, Blattmann P, Thorp TG** (2012) Vascular 4 functioning and the water balance of ripening kiwifruit ( _Actinidia chinensis_ ) berries. J 5 Exp Bot **63** : 1835–1847 6 **Currier HB, Strugger S** (1956) Aniline blue and fluorescence microscopy of 7 callose in bulb scales of _Allium cepa_ L. Protoplasma **45** : 552–559 8 **Da Costa M, Hamann T** (2016) Vascular tissue development. Curr Opin Plant 9 Biol **34** : 118–124 10 **Esau K (1969)** The Phloem. Gebrüder Borntraegger, Berlin 11 **Falchi R, Bonghi C, Drincovich MF, Famiani F, Lara MV, Walker RP,** 12 **Vizzotto G** (2020) Sugar metabolism in stone fruit: source-sink relationships and 13 environmental and agronomical effects. Front Plant Sci **11** : 573982 14 **Grappadelli LC, Morandi B, Manfrini L, O’Connell M** (2019) Apoplasmic 15 and simplasmic phloem unloading mechanisms: do they co-exist in Angeleno plums 16 under demanding environmental conditions? J Plant Physiol **237** : 104–110 17 **Greenspan MD, Shackel KA, Matthews MA** (1994) Developmental changes 18 in the diurnal water budget of the grape berry exposed to water deficits. Plant Cell 19 Environ **17** : 811–820 20 **He K, Niklas KJ, Niinemets Ü, Wang J, Jiao Y, Shi P** (2024) Significant 21 correlation between leaf vein length per unit area and stomatal density: evidence from 22 Red Tip and Chinese photinias. Front Plant Sci **15** : 1365449 23 **Hughes J, McCully ME** (1975) The use of an optical brightener in the study of 24 plant structure. Stain Technol **50** : 319–329 25 **Jensen KH** (2018) Phloem physics: mechanisms, constraints, and perspectives. 26 Curr Opin Plant Biol **43** : 96–100 27 **Jensen KH** (2019) Modeling the hydraulic conductivity of phloem sieve 28 elements. In AM van Bel, B Ehlers, eds, Phloem: Methods and Protocols. Springer, 29 New York, pp 339–344 17 [bioRxiv preprint doi: https://doi.org/10.64898/2026.10.31.762750; this version posted November 2, 2026. The copyright holder for this](https://doi.org/10.64898/2026.10.31.762750) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 1 **Jensen KH, Liesche J, Bohr T, Schulz A** (2012a) Universality of phloem 2 transport in seed plants. Plant Cell Environ **35** : 1065–1076 3 **Jensen KH, Mullendore DL, Holbrook NM, Bohr T, Knoblauch M, Bruus** 4 **H** (2012b) Modeling the hydrodynamics of phloem sieve plates. Front Plant Sci **3** : 151 5 **Jensen KH, Savage JA, Holbrook NM** (2013) Optimal concentration for sugar 6 transport in plants. J R Soc Interface 10: 20130055 7 **Kaneko T, Gould N, Campbell D, Clearwater MJ** (2024) Isohydric stomatal 8 behaviour alters fruit vascular flows and minimizes fruit size reductions in drought 9 stressed ‘Hass’ avocado ( _Persea americana_ Mill.). Ann Bot **133** : 969–982 10 **Kaneko T, Gould N, Campbell D, Snelgar P, Clearwater MJ** (2022) The 11 effect of soil type, fruit load and shaded area on ‘Hass’ avocado ( _Persea americana_ 12 Mill.) water use and crop coefficients. Agric Water Manage 264: 107519 13 **Knoblauch M, Knoblauch J, Mullendore DL, Savage JA, Babst BA,** 14 **Beecher SD, Dodgen AC, Jensen KH, Holbrook NM** (2016) Testing the Münch 15 hypothesis of long distance phloem transport in plants. eLife **5** : e15341 16 **Kondhare KR, Patil AB, Giri AP** (2021) Auxin: an emerging regulator of 17 tuber and storage root development. Plant Sci **306** : 110854 18 **Lang A** (1990) Xylem, phloem and transpiration flows in developing apple 19 fruits. J Exp Bot **41** : 645–651 20 **Liesche J, Bulygin S, Schulz A** (2022) Phloem transport: a historical 21 perspective on the Münch hypothesis and its future challenges. Plant Physiol **189** : 22 1801–1818 23 **Liu X, Robinson PW, Madore MA, Witney GW, Arpaia ML** (1999) Hass' 24 avocado carbohydrate fluctuations. II. Fruit growth and ripening. J Am Soc Hortic Sci 25 **124** : 676–681 26 **Liu Z, Ruonala R, Helariutta Y** (2024) Control of phloem unloading and root 27 development. J Plant Physiol **295** : 154203 18 [bioRxiv preprint doi: https://doi.org/10.64898/2026.10.31.762750; this version posted November 2, 2026. The copyright holder for this](https://doi.org/10.64898/2026.10.31.762750) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 1 **López-Hoyos JH, Henao-Rojas JC, Osorio-Vega NW, Ramírez-Gil JG** 2 (2024) Edaphoclimatic variation and harvest seasonality as determining factors of 3 multidimensional quality in avocado cv. hass grown in the tropics. Heliyon **10** : e33350 4 **Losada JM, He Z, Holbrook NM** (2022) Sieve tube structural variation in 5 _Austrobaileya scandens_ and its significance for lianescence. Plant Cell Environ **45** : 6 2460–2475 7 **Losada JM, Holbrook NM** (2019) Scaling of phloem hydraulic resistance in 8 stems and leaves of the understory angiosperm shrub _Illicium parviflorum_ . Am J Bot 9 **106** : 244–259 10 **Ma S, Li Y, Li X, Sui X, Zhang Z** (2019) Phloem unloading strategies and 11 mechanisms in crop fruits. J Plant Growth Regul **38** : 494–500 12 **Marcelis LFM, Heuvelink E** (2007) Concepts of modelling carbon allocation 13 among plant organs. In Frontis Workshop on Functional-Structural Plant Modelling in 14 Crop Production, pp 103–111 15 **Morandi B, Losciale P, Manfrini L, Zibordi M, Anconelli S, Pierpaoli E,** 16 **Grappadelli LC** (2014) Leaf gas exchanges and water relations affect the daily patterns 17 of fruit growth and vascular flows in Abbé Fétel pear ( _Pyrus communis_ L.) trees. Sci 18 Hortic **178** : 106–113 19 **Morandi B, Manfrini L, Losciale P, Zibordi M, Corelli Grappadelli L** 20 (2010) Changes in vascular and transpiration flows affect the seasonal and daily growth 21 of kiwifruit ( _Actinidia deliciosa_ ) berry. Ann Bot **105** : 913–923 22 **Morandi B, Rieger M, Grappadelli LC** (2007) Vascular flows and 23 transpiration affect peach ( _Prunus persica_ Batsch.) fruit daily growth. J Exp Bot **58** : 24 3941–3947 25 **Mullendore DL, Windt CW, Van As H, Knoblauch M** (2010) Sieve tube 26 geometry in relation to phloem flow. Plant Cell **22** : 579–593 27 **Münch E** (1930) Die Stoffbewegungen in der Pflanze. Gustav Fischer, Jena, 28 Germany 29 **Pleguezuelo CRR, Martínez JRF, Tejero IFG, Ruíz BG, Tarifa DF, Zuazo** 30 **VHD** (2018) Avocado ( _Persea americana_ Mill.) trends in water-saving strategies and 19 [bioRxiv preprint doi: https://doi.org/10.64898/2026.10.31.762750; this version posted November 2, 2026. The copyright holder for this](https://doi.org/10.64898/2026.10.31.762750) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 1 production potential in a Mediterranean climate, the study case of SE Spain: a review. 2 In Water Scarcity and Sustainable Agriculture in a Semiarid Environment, pp 317–346 3 **Poorter H, Niinemets Ü, Poorter L, Wright IJ, Villar R** (2009) Causes and 4 consequences of variation in leaf mass per area (LMA): a meta-analysis. New Phytol 5 **182** : 565–588 6 **Potkay A, Hölttä T, Trugman AT, Fan Y** (2022) Turgor-limited predictions of 7 tree growth, height and metabolic scaling over tree lifespans. Tree Physiol **42** : 229–252 8 **R Core Team** (2023) R: A Language and Environment for Statistical 9 Computing. R Foundation for Statistical Computing, Vienna, Austria 10 **Ræbild A, Anamthawat-Jónsson K, Egertsdotter U, Immanen J, Jensen** 11 **AM, Koutouleas A, Vivian-Smith A** (2024) Polyploidy – a tool in adapting trees to 12 future climate changes? A review of polyploidy in trees. For Ecol Manage **560** : 121767 13 **Ray DM, Savage JA** (2021) Seasonal changes in temperate woody plant 14 phloem anatomy and physiology: implications for long-distance transport. AoB Plants 15 **13** : plab028 16 **Ren Y, Liao S, Xu Y** (2023) An update on sugar allocation and accumulation in 17 fruits. Plant Physiol **193** : 888–899 18 **Rosenberg E** (2022) Three or four levels of hierarchy minimize hydraulic 19 power in leaves with pinnate dendritic venation. J Theor Biol 539: 111061 20 **Savage JA, Beecher SD, Clerx L, Gersony JT, Knoblauch J, Losada JM,** 21 **Holbrook NM** (2017) Maintenance of carbohydrate transport in tall trees. Nat Plants **3** : 22 965–972 23 **Savage JA, Clearwater MJ, Haines DF, Klein T, Mencuccini M, Sevanto S,** 24 **Zhang C** (2016) Allocation, stress tolerance and carbon transport in plants: how does 25 phloem physiology affect plant ecology? Plant Cell Environ **39** : 709–725 26 **Savage JA, Haines DF, Holbrook NM** (2015) The making of giant pumpkins: 27 how selective breeding changed the phloem of _Cucurbita maxima_ from source to sink. 28 Plant Cell Environ **38** : 1543–1554 20 [bioRxiv preprint doi: https://doi.org/10.64898/2026.10.31.762750; this version posted November 2, 2026. The copyright holder for this](https://doi.org/10.64898/2026.10.31.762750) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 1 **Sellier D, Mammeri Y** (2019) Diurnal dynamics of phloem loading: theoretical 2 consequences for transport efficiency and flow characteristics. Tree Physiol **39** : 300– 3 311 4 **Sevanto S** (2018) Drought impacts on phloem transport. Curr Opin Plant Biol 5 **43** : 76–81 6 **Stanfield RC, Forrestel EJ, Elmendorf KE, Bagshaw SB, Bartlett MK** 7 (2024) Phloem anatomy predicts berry sugar accumulation across 13 wine-grape 8 cultivars. Front Plant Sci **15** : 1360381 9 **Thompson MV** (2006) Phloem: the long and the short of it. Trends Plant Sci **11** : 10 26–32 11 **Thompson MV, Holbrook NM** (2003) Application of a single-solute non 12 steady-state phloem model to the study of long-distance assimilate transport. J Theor 13 Biol **220** : 419–455 14 **Thompson MV, Holbrook NM** (2003b) Scaling phloem transport: water 15 potential equilibrium and osmoregulatory flow. Plant Cell Environ **26** : 1561–1577 16 **Torode TA, O’Neill R, Marcus SE, Cornuault V, Pose S, Lauder RP,** 17 **Kračun SK, Rydahl MG, Andersen MCF, Willats WGT, et al** (2018) Branched 18 pectic galactan in phloem-sieve-element cell walls: implications for cell mechanics. 19 Plant Physiol **176** : 1547–1558 20 **Turgeon R** (2010) The role of phloem loading reconsidered. Plant Physiol **152** : 21 1817–1823 22 **Wen S, Neuhaus HE, Cheng J, Bie Z** (2022) Contributions of sugar 23 transporters to crop yield and fruit quality. J Exp Bot **73** : 2275–2289 24 **Whiley AW, Rasmussen TS, Saranah JB, Wolstenholme BN** (1996) Delayed 25 harvest effects on yield, fruit size and starch cycling in avocado ( _Persea americana_ 26 Mill.) in subtropical environments. II. The late-maturing cv. Hass. Sci Hortic **66** : 35–49 27 **Winkler A, Brüggenwirth M, Ngo NS, Knoche M** (2016) Fruit apoplast 28 tension draws xylem water into mature sweet cherries. Sci Hortic **209** : 270–278 21 [bioRxiv preprint doi: https://doi.org/10.64898/2026.10.31.762750; this version posted November 2, 2026. The copyright holder for this](https://doi.org/10.64898/2026.10.31.762750) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 1 **Zamski E** (2017) Anatomical and physiological characteristics of sink cells. In 2 E Zamski, A Schaffer, eds, Photoassimilate Distribution in Plants and Crops: Source– 3 Sink Relationships. CRC Press, Boca Raton, FL, pp 283–310 4 **Zhang XY, Wang XL, Wang XF, Xia GH, Pan QH, Fan RC, Wu FQ, Yu** 5 **XC, et al** (2006) A shift of phloem unloading from symplasmic to apoplasmic pathway 6 is involved in developmental onset of ripening in grape berry. Plant Physiol **142** : 220– 7 232 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 22 [bioRxiv preprint doi: https://doi.org/10.64898/2026.10.31.762750; this version posted November 2, 2026. The copyright holder for this](https://doi.org/10.64898/2026.10.31.762750) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. # 1 FIGURES 24 **Figure 1. Morphology of mature avocado trees during flowering and fruit** 25 **development.** (A) Avocado tree at the stage of full bloom showing mostly inflorescences. 26 (B) Detail of an inflorescence with terminal new vegetative growth (red arrow). (C) 27 Individual flower at anthesis subtended by the pedicel (yellow arrow). (D) Mature tree 28 during fruit growth covered with leaves and showing a second flush of leaves (red 29 arrows). (E) Detailed view of the second flush of leaves (red arrows). (F) Developing 30 avocado pending on the pedicel (yellow arrow). (G) Water content (%) of leaves, flowers, 31 and fruits at different developmental stages, represented by the month of growth (7: July; 32 10: October; 12: December). Letters over bars indicate significant differences (ANOVA 33 and Tukey's post hoc test, _P_ < 0.05). 23 [bioRxiv preprint doi: https://doi.org/10.64898/2026.10.31.762750; this version posted November 2, 2026. The copyright holder for this](https://doi.org/10.64898/2026.10.31.762750) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 **Figure 2. Fruit and pedicel grow during development.** (A) Lenth to width ratio of the 23 flower ovary (orange, Flw) and the developing fruits at months 7 (initial fruit set, pale 24 green), 10 (maximum size attained, light green), and 12 (maturity, dark green). (B) 25 Pedicel diameter at the different developmental stages. Letters over bars indicate 26 significant differences (ANOVA and Tukey's post hoc test, _P_ < 0.05). 24 [bioRxiv preprint doi: https://doi.org/10.64898/2026.10.31.762750; this version posted November 2, 2026. The copyright holder for this](https://doi.org/10.64898/2026.10.31.762750) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 1 **Figure 3. Non-structural carbohydrate (NSC) content in avocado leaves, flowers,** 2 **and developing fruits.** Concentrations are shown for leaves of the previous year (Leaf 0, 3 outline pointing down), the current year (Leaf 1, outline pointing upright), flowers (Flw), 4 and developing fruits in July (Fr07) and December (Fr12). (A) Starch. (B) Sucrose. (C) 5 Fructose. (D) Glucose. Letters over bars indicate significant differences (ANOVA and 6 Tukey's post hoc test, _P_ < 0.05). 7 8 9 10 11 12 13 14 15 16 17 18 19 25 [bioRxiv preprint doi: https://doi.org/10.64898/2026.10.31.762750; this version posted November 2, 2026. The copyright holder for this](https://doi.org/10.64898/2026.10.31.762750) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 1 2 3 **Figure 4. Phloem geometry across leaf vein orders in avocado.** (A) Avocado leaf and 4 close-ups of the different vein orders starting in the petiole (Pet), primary vein (Pri), 5 secondary vein (Sec), tertiary vein (Ter), fourth-order veins (Fou), and fifth-order veins 6 (Fiv). (B) Fluorescence images of individual sieve tube elements (dashed rectangle) 7 stained with aniline blue for callose (yellow) and counterstained with Calcofluor White 8 for cellulose (blue) at the indicated vein orders. (C) Sieve tube element length per vein 9 order. (D) Sieve tube element diameter per vein order. Letters within violin plots indicate 10 significant differences (ANOVA and Tukey's post hoc test, _P_ < 0.05). 11 12 26 [bioRxiv preprint doi: https://doi.org/10.64898/2026.10.31.762750; this version posted November 2, 2026. The copyright holder for this](https://doi.org/10.64898/2026.10.31.762750) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 **Figure 5. Phloem geometry in the petioles and pedicels of avocado flowers and fruits** 25 **during development.** (A) Sieve tube element diameter. (B) Sieve tube element length. 26 Letters within violin plots indicate significant differences (ANOVA and Tukey's post hoc 27 test, _P_ < 0.05). 28 29 30 31 32 27 [bioRxiv preprint doi: https://doi.org/10.64898/2026.10.31.762750; this version posted November 2, 2026. The copyright holder for this](https://doi.org/10.64898/2026.10.31.762750) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 1 2 3 **Figure 6. Sieve pore measurements in the pedicels of avocado flowers and fruits** 4 **during development.** (A) SEM images of sieve plates in the pedicels of flowers (Ped 0), 5 developing fruits in July (Ped 7) and mature fruits in December (Ped 12). (B) Average 6 pore radius per plate at each developmental stage. (C) Average number of pores per sieve 7 plate at each developmental stage. Letters within violin plots represent significant 8 differences (ANOVA and Tukey´s post hoc test at _P_ <0,05). (A) scale bars: 1µm. 9 10 11 28 [bioRxiv preprint doi: https://doi.org/10.64898/2026.10.31.762750; this version posted November 2, 2026. The copyright holder for this](https://doi.org/10.64898/2026.10.31.762750) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 1 2 **Figure 7. Sieve tube element number per cross-section in petioles and pedicels** 3 **during development.** Tissues analyzed include petioles (Pet) and pedicels of flowers 4 (Ped 0), developing fruits in July (Ped 7), and mature fruits (Ped 12). (A) Processed 5 images stained with calcofluor White for cellulose and pectins (blue color) with isolated 6 phloem at each developmental stage. (B) Total number of tubes per cross-section (left Y7 axis) and sieve tube element density (right Y-axis) at each developmental stage. 8 9 29 [bioRxiv preprint doi: https://doi.org/10.64898/2026.10.31.762750; this version posted November 2, 2026. The copyright holder for this](https://doi.org/10.64898/2026.10.31.762750) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 1 2 3 **Figure 8. Sieve tube conductivity from source to sink tissues over avocado** 4 **development.** Conductivity was calculated from phloem in source tissues—fifth-order 5 (Fiv), fourth-order (Fou), tertiary (Ter), secondary (Sec), and primary veins (Pri), and 6 petiole (Pet)—to sink tissues, including the pedicels of flowers (Ped 0), developing fruits 7 in July (Ped 7), and mature fruits (Ped 12). 30 [bioRxiv preprint doi: https://doi.org/10.64898/2026.10.31.762750; this version posted November 2, 2026. The copyright holder for this](https://doi.org/10.64898/2026.10.31.762750) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 1 2 **Figure 9. Modeling phloem transport in avocado from source petioles to sink** 3 **pedicels.** Tissues modeled include sink pedicels of flowers (Ped 0. orange), developing 4 fruits (Ped 7, light green), and mature fruits (Ped 12, dark green). (A) Conductance 5 (ribbons) and conductivity (circles) during fruit development show similar trend of 6 increase. (B) Calculated pressure drop of the phloem considering a fixed sap flow (Q) and 7 a predicted phloem operation within the 0.1 m–1.0 m length range (triangles and squares). 8 (C) Calculated flow rate Q considering a fixed pressure drop of 0.1 MPa. within the 0.1 9 m–1.0 m length range (triangles and squares), compared with experimental values 10 (ribbons). 11 31