Conceptually, the model was divided into three components: plant water status, gas exchange and phloem transport (Figure 2). The plant water status component calculated the leaf mesophyll and xylem water potentials (Ψ_m and Ψ_x) from an input soil water potential (Ψ_s), gas exchange and hydraulics. The gas exchange component calculated photosynthesis from the stomatal aperture, which was determined from the mesophyll water status (Ψ_m). The phloem transport component calculated a source – sink phloem sucrose concentration and pressure gradient generated from the Michaelis-Menten kinetics of loading/unloading sucrose and the water stress experienced by the phloem (Ψ_x or Ψ_s).

To simplify the model, we made the following assumptions:

We separated the plant into two hydraulic elements, the mesophyll of a single leaf and the root-to-leaf xylem network, to capture the protective effect of vulnerability segmentation (Tyree and Ewers, 1991). The leaf accounts for at least 30% of whole-plant hydraulic resistance, with the mesophyll tissue accounting for about half of this resistance (Sack and Holbrook, 2006; Scoffoni et al., 2011). This generates a strong water potential gradient across the mesophyll that buffers the xylem and phloem against water stress. We adapted the water balance equations from Sperry et al. (1998) to calculate the mesophyll and xylem water volume at each timestep:

where W is the water volume, Ψ is the water potential, and K is the hydraulic conductance of the mesophyll (subscript m) and xylem (subscript x). The VPD and Ψ_s are the environmental parameters, the vapor pressure deficit and soil water potential, respectively; g_s is the stomatal conductance, and LA is the area of a single leaf (see Table 1 for constant parameter values and Table 2 for the parameters varied across simulations).

The mesophyll water volume is converted to a relative water content by dividing by the maximum water volume (RWCm=WmVsat,m), and then to a water potential through the pressure-volume relationships:

where π_o, π_tlp, a_f, and ε are the leaf pressure-volume curve parameters of osmotic potential at full hydration turgor loss point, apoplasmic fraction, and cell wall modulus of elasticity, respectively (Bartlett et al., 2012). We considered the effects of mesophyll sugar content on water relations to be outside the scope of this study, and, thus, we assumed here that π_o was constant.

Water flow through the mesophyll was determined by integrating the mesophyll vulnerability curve:

where K_max,m is the maximum hydraulic conductance of the mesophyll, normalized by leaf area, α is a shape parameter, and Ψ_50,m is the mesophyll water potential at which 50% of conductance is lost. To simplify these calculations, we used a sufficiently negative xylem Ψ₅₀ value (−2 MPa) to assume K_x was constant and equal to the maximum xylem conductance (K_max,x) over the range of xylem water potentials in these simulations.

We calculated g_s from the assumption that mesophyll water stress induced stomatal closure:

where g_max is the maximum stomatal conductance, ags is the shape parameter for this relationship, and g_{s 50} is the mesophyll water potential inducing 50% stomatal closure. Photosynthesis (A_net) was then calculated from g_s based upon an extrapolation of the original Farquhar equations (Farquhar et al., 1980; Bartlett et al., 2019).

At the beginning of each timestep, Ψ_m was calculated from the change in mesophyll volume over the previous timestep (Eqs 2a, 3). Mesophyll water flow was then calculated by integrating the mesophyll vulnerability curve over water potential, bounded by the new Ψ_m and the Ψ_x from the previous timestep (Eqs 2, 3). Stomatal conductance (g_s) was then calculated from the new Ψ_m (Eq. 5), and Ψ_x was updated from the new mesophyll water flow:

where t indicates the current timestep and was used to determine the new xylem flow K_max,x(Ψ_s−Ψ_x) (Eq. 2b). The new mesophyll flow and g_s values were then supplied to Eq. 2a to update the mesophyll volume for the next timestep. We ran the model at a 1s timestep over 12 h simulations to achieve steady-state solutions, which were reported as the model results.

The mass of sucrose in the mesophyll (S_m) was increased by photosynthesis and reduced by loading into the phloem companion cells via sucrose transporters:

where L_sucL is the loading rate per unit leaf area. To simplify the model, we assumed all the sugar produced by photosynthesis is exported into the apoplasmic space surrounding the sieve element/companion cell complexes of the loading zone, and thus available for phloem loading. For loading (subscript i = L) or unloading (subscript i = U), L_suc,i was calculated from the Michalis-Menten formalism for enzyme kinetics:

where V_max,i is the maximum loading or unloading rate of the sucrose transporters, normalized by leaf area, and M_S is a shape parameter capturing transporter affinity for sucrose. C is either C_m, the sucrose concentration in the mesophyll outside the loading zone, or C_pU, the sucrose concentration in the unloading phloem zone. W_m is the mesophyll water volume (Eq. 2a), which changes the concentration of sucrose available for loading in Eq. 8b.

Loading increased the mass of sucrose in the phloem loading zone (S_p,L), which was transported to the unloading zone:

where E_s is the mass flow rate of sucrose transport (g s^–1). E_s was calculated from the volumetric flow rate of the phloem sap (F_p; m³ s^–1) and sucrose concentration in the loading zone (C_p,L; mol m^–3):

where mm is the molar mass of sucrose and W_p is the maximum water volume in the loading zone. Thompson and Holbrook (2003a) found that accounting for diurnal changes in phloem volume did not substantially impact the flux of sucrose through the transport pipeline. Thus, we made the simplifying assumption that W_p was constant at the maximum loading zone phloem volume, so that C_pL only varied with the mass of sucrose.

The sucrose concentration determined the osmotic potential (π_i) in the loading (L) or unloading zone (U):

where Rc is the gas constant and T is the temperature within the phloem loading area. Following Thompson and Holbrook (2003b), we made the simplifying assumption that the water potential of the phloem loading zone equilibrates with the xylem water potential (Ψ_x) when the phloem is at steady state. Further, we estimate that the anatomical dimensions, osmotic strength, and enhanced permeability due to aquaporins (Muries et al., 2013; Stanfield et al., 2017) of our loading zone has an RF (axial to radial resistivity) value >> 1 (see Thompson and Holbrook, 2003b), which equates to phloem in water potential equilibrium with the xylem. The turgor pressure in the loading or unloading zone (P_i) was then determined from the water and osmotic potentials:

Noting that for determining the phloem pressure in the unloading zone, Ψ_s is used. The volumetric phloem flow rate F_p (m³ s^–1) was calculated from the pressure difference between the loading (P_L) and unloading (P_U) zones and the hydraulic resistance of the phloem (R_conduit):

which increased with the phloem sucrose concentration due to viscosity (v) effects (see Appendix A), and integrated into a resistance formula:

where Rp is the resistance coefficient of the single phloem conduit, or sieve tube (see Table 2).

The velocity of sap flow (V_s) was calculated by normalizing volumetric flow by conduit area:

where r is the radius of the conduit.

The same processes then take place when the sucrose reaches the unloading zone, where unloading is determined from:

We made the simplifying assumptions that the source and sink area (LA) and the phloem volume in the loading and unloading zones (W_p) are equal.

Finally, we quantified sugar export (S_tot) as the cumulative mass of sucrose unloaded over the simulation, as:

The model was implemented using MATLAB R2020a (9.8.0) (Mathworks, Inc., Natick, MA, United States). Access to the code may be obtained through doi: 10.5281/zenodo.5907490.

We tested the impacts of water stress and phloem anatomy and transporter kinetics on phloem function by varying the (1) soil water potential (Ψ_s), (2) phloem structural resistance (Rp), (3) phloem volume (Wp), (4) maximum sucrose loading rate (V_maxL), and (5) the shape parameter for the relationship between loading rate (L_sucL) and turgor (β) across simulations (see Table 2 for parameter values). We parameterized a range in soil water stress by varying Ψ_s from well-watered to droughted values (−0.001 to −1 MPa). We used the published sieve element dimensions for stems from 5 active loading species to produce reasonable estimates of Rp (Figure 1; see model assumption #7). Data for source phloem volume is scarce, so we initially estimated Wp by multiplying the mean sieve element area for Populus tremula x alba leaf (Carvalho et al., 2017) by the total vein length in a grapevine leaf, as a representative active loader (Pagano et al., 2016). However, this estimation only accounted for 0.002% of total volume in our hypothetical leaf, while the phloem has been estimated to account for up to 0.4% of leaf volume (Sjolund, 1997). Thus, we parameterized the model with phloem volumes ranging from 0.002 to 0.8% of leaf volume to capture a wide range of potential parameter space. We compared the cumulative sucrose export (S_tot), phloem loading zone concentration (C_l), pressure (P_l), phloem viscosity (v) and velocity (V_s) across simulations.

Phloem resistance and volume had strong interactive effects on sugar export, with small volumes exacerbating viscosity limitations for high resistances (see “Results,” Figure 3). Thus, we were concerned that incorrect assumptions about these parameter combinations could overestimate viscosity limitations under drought. We used our first simulations to identify the phloem volumes that (1) maximized sugar export with (2) viable phloem pressures under drought (Ψ_s = −1 MPa) for each Rp value. We used these values to parameterize the rest of the simulations, to evaluate the impacts of transporter kinetics on viscosity limitation under the most favorable anatomical parameterizations.

Next, we parameterized variation in maximum loading rate (V_maxL) by compiling published values for low (Arabidopsis), intermediate (tobacco), and high (wheat) rates (Weise et al., 2000; Borstlap and Schuurmans, 2004; Liesche and Schulz, 2013). We used the intermediate value from tobacco as our default unless otherwise noted (Table 3). We also tested whether regulating loading rates in response to pressure would protect the phloem from viscosity limitations, by reducing loading before reaching excessive concentrations and pressures. We used an exponential decay function to reduce the loading rate as a function of loading zone pressure:

and increased the shape parameter (β) from 0 to 10 to test the impacts of loading downregulation. We first varied β and V_maxL independently, and then together, to identify the combinations that (1) maximized sucrose export and (2) obtained viable loading zone pressures (see model assumptions #8) for different resistances (R2 and R4) and soil moisture scenarios (Ψ_s = −0.001 and −1 MPa).

The importance of viscosity limitations on sugar transport depended strongly on phloem anatomy (Figure 4). Simulations either reached stable, equilibrium sucrose concentrations over the course of the 12-h model runs (i.e., steady-state), or failed to converge on a stable concentration due to excessive viscosity (i.e., non-steady-state) under high water stress and/or sieve tube structural resistances and small phloem volumes. For example, a simulation with moderate structural resistance (R2) reached steady-state and achieved a stable loading zone concentration, pressure, total resistance, and flow rate under dry conditions (ψ_s = −0.001 and −1 MPa) (Figure 4, yellow lines). By contrast, a simulation with ∼25% greater structural resistance still reached steady-state in wet soil (Figure 4, blue lines), but failed to equilibrate under water stress and phloem concentration and pressure increased without limit (Figures 4A,B, red lines). In these non-steady-state simulations, phloem viscosity and total resistance remained too high for the loading zone pressure to overcome (Figure 4C, red line), even as concentrations increased, preventing phloem flow (Figure 4D, red line). In these scenarios, sugar export only proceeded for a small fraction of the 12-h simulation.

Coordinating phloem volume with pathway resistance alleviated runaway viscosity and avoided non-steady-state transport failure. There is little data available to parameterize phloem volume and resistance in leaves for active loading species, so we simply varied these parameters widely (i.e., by three orders of magnitude for phloem volume; see Methods) to evaluate the impacts on sucrose export (Figures 3A,B), phloem pressure (P_l; Figures 3C,D) and viscosity (v; Figures 3E,F). In the lower resistance pathways (R1 and R2), all phloem volumes allowed sucrose export to proceed without runaway viscosity, while the smallest volumes produced the greatest sucrose output, under both wet and dry conditions (Ψ_s = −0.001 and −1 MPa). However, for the intermediate (R3) and high (R4 and R5) phloem resistances, phloem pressure and viscosity reached intractable values at lower volumes; this effect was exacerbated under drier soils. Overall, larger phloem volumes were optimal for sucrose export in more resistive pathways and drier conditions. We also used the volumes that optimized export under drought (white asterisks) to parameterize volume for each resistance in the subsequent simulations, to avoid overestimating the importance of viscosity limitations by making incorrect assumptions about phloem anatomy (i.e., W_p = 3.57e^–11 m³ for R1 and R2, 1.04e^–8 m³ for R3, and 1.37e^–8 m³ for R4 and R5; see “Materials and Methods” and Table 2).

Sugar export was constant over the soil water potentials tested for structural resistances below thresholds for viscosity limitation (Figure 5A, R1 and R2), and declined above these thresholds (Figure 5A, R3–R5). Below these thresholds, loading zone concentration and turgor increased until flow reached a steady-state equilibrium (see Figure 4), where the mass of sucrose imported into the phloem equaled the mass exported. Here, we assumed transport was not constrained by unloading and that all sucrose that entered the unloading area was immediately exported to sinks (i.e., V_maxU >> V_maxL; see “Materials and Methods”: assumption #2). Thus, sucrose export in these lower resistance simulations (R1 and R2) were independent of water stress and phloem pressure. Drier conditions increased sap viscosity and reduced phloem velocity by changing the total resistance and source to sink pressure gradient (Figures 5B,C). For the least resistant phloem (Figure 5D, R1), phloem pressure slightly decreased with soil water potential, because the xylem water potential was decreasing at a greater rate than phloem osmotic potential (e.g., Supplementary Figure 3, green line).

In contrast, the simulations with the most resistant sieve tubes (R3–R5) not only declined in their overall sucrose output but varied in their response to the soil dry-down (Figure 5A). The decline in sucrose output was attributed to a viscosity limitation, as the higher resistance scenarios had both higher overall viscosity, and increased viscosity under higher drought stress (Figure 5B). As with the lower resistance pathways, velocity declined with soil water potential, but did so more rapidly (Figure 5C). Finally, phloem pressures in the higher resistance pathways were well above empirically measured values of 0.6–2.4 MPa (Figure 5D, red dashed lines). Of note was the intermediate resistance (R3) which showed increased phloem pressure with declining soil water, as a result of phloem osmotic potential increasing faster than xylem water potential (Supplementary Figure 3, red line).

Although increasing phloem volume helped alleviate runaway viscosity, higher resistance pathways where still under the effects of viscosity limitation, which caused phloem pressures to be unrealistic. Thus, we hypothesized maximum loading rates to be modulated according to these anatomical constraints, by either downregulating V_maxL constitutively (Figures 6A,C) or inducibly as a function of pressure (Figures 6C,D). Reducing the maximum loading rate by half (V_maxL) reduced sucrose output to sinks (Figure 6A) in comparison to the non-regulated loading, but lowered phloem pressures to reasonable bounds (Figure 6C, red dashed lines). However, using Eq. 18 to downregulate phloem loading as a function of pressure improved overall sucrose output over simply reducing V_maxL while maintaining viable phloem pressures over most soil water conditions (Figures 6B,D). Comparing wet soils (Ψ_s = −0.001MPa) between unregulated loading (Figure 5A) to pressure regulated loading (Figure 6B), sucrose output dropped by 30.6, 27.8, and 2.36% for R1, R3, and R5 resistances, respectively (β = 0.6). However, under dry soils (Ψ_s = −0.001 MPa), sucrose output dropped by 22.9 and 8.5% for R1 and R3, respectively, but for the highest resistance (R5), improved by 3.2%. This led to the hypothesis that pressure regulated loading may be optimized according to both sieve tube anatomical traits (e.g., Rp), environmental conditions (e.g., Ψ_s) or maximum loading rates (e.g., V_maxL).

We next sought to determine how the transporter kinetics that maximize sucrose export while maintaining viable phloem pressures depend on pathway resistance and soil moisture. We first identified the range of viable β values for each resistance and soil moisture scenario (i.e., Ψ_s = −0.001 and −1 MPa), assuming a constant V_maxL (i.e., for N. tabacum) (Supplementary Figure 4). We then varied V_maxL from 1.30e^–8 (i.e., Arabidopsis) to 2.30e^–6 mol m^–2 s^–1 (i.e., wheat) to identify the combinations of V_maxL and β that maximized export for each scenario (Figure 7).

Coordination between loading downregulation and the maximum loading rate strongly benefitted sucrose export (Figure 7). This simulation showed sugar export for the β and V_maxL combinations that produced viable phloem pressures (i.e., 0.6–2.4 MPa). For low-resistance phloem (R2) under wet conditions, sucrose export was maximized at 19.07 mg by V_maxL = 1.79e^–6 mol m^–2 s^–1 and β = 0.79 (Figure 7A). Drought reduced the maximum output at this Rp to 13.6 mg (i.e., a 29% reduction) (Figure 7C), which occurred at a higher V_maxL (2.22e^–6 mol m^–2 s^–1) with stronger pressure-coupled downregulation (β = 1.14); in comparison to default V_maxL, the expanded V_maxL improved export by 50 and 28% for the wet and dry conditions, respectively, at this resistance.

Increasing resistance (R4) in wet conditions substantially reduced maximum output to only 4.18 mg (Figure 7B); this corresponded to a lower V_maxL and stronger downregulation than for R2 (V_maxL = 1.68e^–6 mol m^–2 s^–1 and a β = 1.33). Water stress reduced maximum output to 2.98 mg (a 29% reduction) (Figure 7D) and this corresponded to the strongest downregulation across scenarios (V_maxL = 1.72e^–6 mol m^–2 s^–1 and β = 1.61); this improved sucrose output by 27 and 33% over default V_maxL values for wet and dry conditions, respectively. Overall, when comparing the impacts of pressure regulated loading (Figure 7) vs. non-regulated loading (Figure 5), we found that non-water stressed, low resistant pathways and water stressed, high resistant pathways benefit the most from a pressure regulated loading mechanism (Table 4).

Our study provides theoretical support for our hypothesis that loading regulation is adaptive for phloem function. Pressure-regulated loading prevented the excessive buildup of sugars that causes viscosity limitations, reducing the impacts of water stress on sugar transport for highly resistive phloem (Figures 6, 7). Regulated loading also benefitted less resistive pathways by allowing for a higher V_maxL and, thus, greater sugar export, than in the absence of regulation (Table 4). These model findings are bolstered by empirical evidence supporting pressure-regulated behavior and the discovery of molecular mechanisms that could dynamically adjust sucrose loading during drought.

Previous experiments have shown that loading rates may be adjusted as a function of turgor and/or sucrose concentration during osmotic stress. For example, exposing tissues to membrane-impermeable sugars (i.e., sorbitol or mannitol) reduces the apoplasmic water potential outside the phloem, drawing out water and reducing phloem turgor (Smith and Milburn, 1980; Patrick, 1994; Bell and Leigh, 1996). These treatments showed that a decline in phloem pressure upregulated sucrose loading in castor bean leaf discs (Smith and Milburn, 1980) and sucrose unloading in bean seed coats (Patrick, 1994). Conversely, reducing external mannitol concentrations increased phloem turgor and decreased unloading rates in beet root disks (Bell and Leigh, 1996). Supporting loading downregulation, isolated vesicles from beet leaf disks exposed to increasing sucrose concentrations showed declines in loading rate in a concentration-dependent manner (Chiou and Bush, 1998). This downregulation also corresponded to a reduction in sucrose symporter transcript abundance with increasing sucrose concentrations, causing maximum loading rate (V_max) to decline by up to 74% compared to controls which were not subjected to increased sucrose concentrations.

The empirical and modeling results suggest that plants would benefit from the ability to exhibit a range of loading rates, while the directionality of loading responses would depend on internal phloem conditions (e.g., anatomy or viscosity). For example, the model predicted that more resistive phloem would require stronger downregulation to optimize sugar transport under drought than less resistive pathways (Figures 7B,D; R4 vs. Figures 7A,C; R2). Weaker downregulation would allow excessive sucrose concentrations to build in the phloem, causing pressure and viscosity to rise and flow rates and export to decline (Figure 6B; R3–R5). Thus, upregulation could be beneficial during drought if resistance is below thresholds for viscosity limitation, allowing greater osmotic strength to compensate for the lowered xylem water potentials.

Changes in the regulation of sucrose transporter activity may be the primary cause of loading up- or downregulation and several mechanisms could allow these changes to occur in response to drought. Alteration in phloem turgor have been hypothesized to trigger a hormonal signaling cascade that alters transporter expression or post-translational modification (Patrick et al., 2001). This has been evidenced by water stress upregulating most (but not all) sucrose transporter subfamilies involved in phloem loading in some species (Medici et al., 2014; Xu et al., 2018). For example, Arabidopsis, soybean, barley, rice, wheat, and maize saw an overall upregulation of phloem loading SUT expression, but potato and tomato did not. However, phloem exudates from these studies indicate that sucrose concentrations decreased under water stress for all species, which contradicts the hypothesis that SUT transcript abundance alone drives differential loading rates but is consistent with our model results for differential control of V_maxL and pressure down-regulation of loading (Figure 7). Alternatively, post-translational modification of transporter proteins may regulate the rate of sugar movement past the phloem membrane (reviewed by Liesche et al., 2011a). For instance, ubiquitination of these proteins increases their degradation rate in the plasma membrane, while phosphorylation increases their affinity for sucrose in plants exposed to differing light (Xu et al., 2020) or drought (Ma et al., 2019). Interestingly, in high light conditions, photosynthesis and SUC2 phosphorylation were significantly increased, while SUC2 transcript remained stable (Xu et al., 2020). This implies that SUC/SUT transporter regulation is complex and under multiple controls. Future experiments may incorporate a variety of progressively intensified abiotic challenges (e.g., high light intensity, drought) and track both expression levels and post-translational modifications of SUTs/SUCs; this would help determine if thresholds of sucrose transporter upregulation exist, beyond which downregulation occurs to prevent viscosity limitation.

The current work focuses on the role of regulating loading proteins at the companion cell/sieve element interface. However, there is emerging evidence that phloem loading may be controlled in the pre-phloem pathway (Liesche, 2017), which would expand the applicability of studying viscosity limitations past strict apoplasmic loaders; for example, passive loaders or species with alternating loading types (i.e., English Oak; Liesche, 2017). Understanding how loading regulation varies between species may also help to elucidate carbon allocation patterns in response to stress (Savage et al., 2016).

For example, poplar (Populus trichocarpa) shows regulation of sucrose in the pre-phloem pathway through differential expression of tonoplast sucrose transporters in mesophyll cells (Payyavula et al., 2011). Under drought stress, these transporters were downregulated, stem growth reduced, and leaf sugar accumulation increased (Frost et al., 2012), signaling that phloem export from the leaf was diminished (Liesche, 2017). It could be hypothesized that if the Populus phloem was on the verge of viscosity limitation, sucrose needed to be withheld from entering the phloem loading pathway by being sequestered in vacuoles to prevent viscosity build-ups. Alternatively, other species such as beech (Fagus sylvatica) did not show reduced carbon export from the leaf during drought, but instead saw a reduction in photosynthesis (Hommel et al., 2016; Liesche, 2017). Beech has been identified as a passive symplasmic loader (Rennie and Turgeon, 2009), which may make it more reliant on adjustments of photosynthesis to control proper sucrose gradients in the phloem. Future work will need to identify sucrose transporter proteins in both the pre-phloem and phloem pathways to understand what role loading regulation plays in supporting efficient phloem transport across multiple plant taxa with differing loading strategies.

Pressure-based loading improved sucrose export to sinks over non-regulated scenarios by up to 3.7x, while the lowest resistance pathway (R1) improved loading over the highest (R5) by 38.7x (Table 4). This might imply that plants with higher growth rates correlate to a lower resistance phloem pathway. However, a recent meta-analysis on phloem anatomical traits did not find a significant trend between growth rates and sieve tube resistance (Liesche et al., 2017) but did find increased variability in taxonomic groups that actively load. Potentially, loading regulation allows for higher pathway resistances to achieve similar levels of export to sinks in comparison to non-regulated pathways, which would make anatomical characters less constrained.

Another interesting interaction in phloem anatomy that was highlighted here was the relationship between phloem volume and resistance. In our simulations, export was maximized at phloem volumes that were small enough for rapid loading, but large enough to avoid viscosity limitations (Figure 3). Further, optimal volumes were larger for more resistant pathways. The relationship between phloem volume and resistance would depend on the underlying traits; for example, doubling conduit radius would increase phloem volume four-fold and reduce resistance 16-fold, while doubling the number of parallel conduits per unit area would decrease resistance and increase volume linearly (Hölttä et al., 2009; Jensen, 2019). This flexibility suggests it is highly plausible for natural or artificial selection to achieve optimal coordination between resistance and volume, though more work is needed to determine whether the inverse relationships between volume and resistance would prevent highly resistive pathways from achieving optimal volumes.

Previous studies have used genetic engineering to upregulate the expression of phloem loading sucrose transporters in pea (Pisum sativum; Lu et al., 2020), Arabidopsis (Dasgupta et al., 2014), potato (Solanum tuberosum, Leggewie et al., 2003), and rice (Oryza sativa; Wang et al., 2015). Although these transformations increased loading rates (Lu et al., 2020), the impacts on growth and viability varied by species. One successful example from pea plants saw the upregulation of PsSut1 which increased sucrose concentrations in the phloem exudate and significantly increased biomass and yield (Lu et al., 2020). The authors hypothesized that upregulating SUT1 enhanced both loading and unloading in the developing seeds, which was key to making this transformation successful (Lu et al., 2020). Similarly, we found that reducing sink limitations would minimize the buildup of sucrose that would encourage viscosity limitation (Supplementary Figure 1). Our study also suggests that lowering phloem pathway resistance in coordination with increasing maximum loading rate would lower pressure induced downregulation and increase sucrose output substantially (Figure 7). This could be achieved by screening potential lines of SUT transformation for sieve tube anatomical characteristics that lower resistance, such as sieve element length, diameter, and sieve plate porosity (Stanfield et al., 2019). While sieve tube anatomy may be a difficult genetic target, reducing plant height (dwarfing) could be a simple method to reduce pathlength resistance (e.g., Qiao and Zhao, 2011) to determine the interacting effects of SUT upregulation, lowered pathway resistance and yield. Further, elucidating the mechanisms that generate turgor-dependent signaling cascades that modify sucrose transporter expression (as suggested by Patrick et al., 2001) could allow genetic engineering approaches to optimally target maximum loading rates.