General Comments:

The manuscript added plant hydraulics' influence on ecosystem carbon and water fluxes into the LandscapeDNDC model. The new model was calibrated and evaluated for a pine plantation in Israel. The manuscript presented the multiple pathways of plant water stress (stomatal, non-stomatal, leaf shedding, and sapwood loss) and their importance in simulating drought impacts. Overall, I like the idea of organizing plant water stress as a sequential physiological response triggered by plant water potential, as summarized by many ecophysiological studies cited in the paper. However, I feel the modeling approach presented is too empirical to make the added processes generalizable and useful (maybe constrained by the model structure of LandscapeDNDC). Particularly, the only real evaluation of the plant hydrodynamic module is pre-dawn water potential, which is highly determined by soil hydraulics instead of plant hydrodynamics. In addition, there are various modeling efforts (mostly in the context of tropical forests) that have implemented sequential responses, which are not acknowledged. Altogether, the essentiality of the added module is not well highlighted and these reduce the significance and novelty of the study.

Specific Comments:

1. The introduction claims that plant hydraulic processes are not represented in a consistent way in ecosystem modeling (Line 50-55, Line 85-90, etc.). This is not true. Christofferson et al. 2016 and Xu et al. 2016 have both fully integrated plant hydrodynamics with plant physiology in demography-explicit ecosystem modeling. In particular, Xu et al. 2016 implemented stomatal, non-stomatal (through a reduction in carboxylation capacity), and phenological responses to drought. Meanwhile, explicit tracking of sapwood dynamics is indeed rare. Most of the modeling practices implicitly include the reduction of sapwood fraction through a reduction in conductivity.

Christoffersen, B. O., Gloor, M., Fauset, S., Fyllas, N. M., Galbraith, D. R., Baker, T. R., Rowland, L., Fisher, R. A., Binks, O. J., Sevanto, S. A., Xu, C., Jansen, S., Choat, B., Mencuccini, M., McDowell, N. G., Meir, P., Kruijt, B., Rowland, L., Fisher, R. A., … Meir, P. (2016). Linking hydraulic traits to tropical forest function in a size-structured and trait-driven model (TFS v.1-Hydro). Geoscientific Model Development, 9(11), 4227–4255. https://doi.org/10.5194/gmd-9-4227-2016

Xu, X., Medvigy, D., Powers, J. S., Becknell, J. M., & Guan, K. (2016). Diversity in plant hydraulic traits explains seasonal and inter-annual variations of vegetation dynamics in seasonally dry tropical forests. The New Phytologist, 212(1), 80–95. https://doi.org/10.1111/nph.14009

2. Many of the hypotheses (Line 108-114) are not appropriate to be answered by a modeling study. For instance, testing the first hypothesis on VPD limitation on stomatal conductance would require physiological observations. If one codes in VPD effect in the model, you will surely see VPD effects independent from soil moisture. Furthermore, isn't the VPD effect already known as early as the Leuning or Collatz stomatal model (if not earlier)... Similarly, the second and third hypotheses need some more thought. One way to postulate useful hypotheses in a modeling study is to test whether a new process shifts model behavior (and under what circumferences) and make model results closer to observations. Hypothesis iv reads better and more interesting compared with other ones.

3. I have a hard time understanding and interpreting psi_dehydration (line 250-255). First, shouldn't psi_xylem be equal to psi_root minus sapflow/transpiration divided by root-to-xylem resistance? I am not certain about the motivation for defining psi_dehydration. Second, psi_dehydration is the average potential gradient between canopy and roots, weighted by foliar biomass, in the past k days. Which is then used to calculate instantaneous canopy water potential later in equation (6a). The mismatch of time-scale (k-day average vs instantaneous) is really confusing and I am confused about the downregulation of the gradient by foliar biomass. I have some other concerns on the hydraulic physiology of the model specified later.

4. I have two main concerns over the results and discussions. First, there are a lack of evaluation/benchmark of the new plant hydraulic module. The only evaluation is Fig. 3a, which shows the model can predict seasonality and diurnal cycle of water potential. However, the model seems to significantly overestimate water potential in the dry season of 2013. I understand hydraulic measurements might be rare to get. Some more discussions on modeled hydrodynamics would be helpful.  For example, what is the diurnal cycle of transpiration/sapflow (I think these are the other hydrodynamic variables available at the site?) compared with the observed sapflow in the wet season and dry season respectively. Second, there is a lack of comparison for models with/without the new additions (except for Fig. 5 when comparing NSL on and off). It is not clear to me how the new additions are essential to correctly model transpiration, GPP, etc. For example, if calibrating the original LandscapeDNDC with MCMC, could the calibrated model capture seasonality in transpiration and GPP? 

Technical Comments:

Line 58, 'Already' is excessive here

Line, 75-80, As I mentioned above, there have been many models that simulate plant hydraulics at stand-scale, including but not limited to ED2 (Xu et al. 2016), ELM-FATES (Christofferson et al. 2016), CLM (Kennedy et al. 2019), JULES (Eller et al. 2018), and ORCHIDEE (Yao et al.2022). There are challenges while much progress has been made already

Line 120-125 What would be the soil water potential for the wilting point? This might be related to the 'disconnect' water potential.

Line 143-145, given there is an EC tower as well. I wonder whether the sapflow-based transpiration has been compared with tower-based ET?

Line 169, is the soil carbon/nitrogen module relevant here? If not, it can be removed.

Line 205-210, I am confused by the variable RPMIN and the calculation of rp. What do they represent physically? In addition, both RPMIN and krc_rel have a unit of mmol/m2/s/MPa, so how could their quotient also have a unit of mmol/m2/s/MPa?

Line. 215-220. So, wdef is some kind of magical residual soil water pool that plants can access in the dry season? This means the hydrological budget of the model is not conserved. How negative wdef can get?

Line 279 'hydraulic vulnerability cure' --> hydraulic vulnerability curve

Line 319, Fig. 1, Soil evaporation is missing in the figure. Is it important in the ecosystem?

Line 345, Fig. 2, the simulated GPP is biased low in 2012-2013 dry season. What could be the potential explanations?

Line 385, Fig. 5, It is great to show model behavior difference. A further question is which one is closer to observations. Is there any way to benchmark these two curves with observations? Maybe you can plot and contrast GPP vs soil moisture for the two simulations as well as observations?

Line 401, Fig. 6, very interesting figure and I really like the implementation of sapwood turnover and growth. Just curious does the sapwood area increment match the observed tree basal area growth at the site?

Line. 470-480, Prieto et al. 2012 has discussed about asymmetric root-soil hydraulic conductivity, which might give rise to the disconnect water potential.

Prieto, I., Armas, C., & Pugnaire, F. I. (2012). Water release through plant roots: new insights into its consequences at the plant and ecosystem level. New Phytologist, 193(4), 830–841. https://doi.org/10.1111/J.1469-8137.2011.04039.X

This manuscript presents and evaluates new developments made on the LandscapeDNDC modelling framework, focused on improving the realism and model performance under drought stress, using data from an extremely dry Aleppo pine plantation. The design of model modifications is sound and addresses several issues that are at the forefront of current modeling efforts in the community. The resulting model seems to perform appropriately, at least according to the observed data sets available. Furthermore, the discussion of the importance of the different processes and representation is interesting. I like Fig. 1 and how the manuscript contribution is framed, in general. My only main concert, however, is on the way objectives are stated, which in my opinion is a bit odd. In particular, I don’t think the questions targeted are those stated. For example, question (i) has an obvious answer, yes, as it is only a matter of model design. Actually, there are other models that separate the influence of soil water potential and leaf water potential on transpiration/photosynthesis. Is the authors’ objective to gain knowledge of the importance of the different processes in the ecophysiology of plants? or to be able to successfully represent those processes in a model framework? and evaluate the sensitivity of model predictions to their representation? I believe the authors seek the second objective, but the current text seems to navigate between both and is not clear in this respect. In my opinion, by adding more clarity in the end of the Introduction, and in the Discussion section, the overall manuscript would improve in usefulness. I have a suggestion related to this. Besides evaluating the model performance with observed data, I suggest the authors to more straightforwardly compare the effect of including the different modifications (hydraulic model, NSL, defoliation) one by one, as done for NSL. With these comparisons, the reader will understand the importance of considering these processes, in the LandscapeDNDC model framework or others.

Minor comments

L47 – Non-stomatal limitations to photosynthesis.

L154 – What about precipitation data? Did came from a gauge in the same tower?

L207 – Please add more details on the formulation of the cost function (xi: ξ).

L209 – is krc_rel a relative root-to-canopy hydraulic conductance or an absolute one with explicit units? Not clear.

L215 – Not clear how gmin relates to wdef. Do you mean that gmin increases wdef progressively, once stomata are closed?

L228 – Are further reductions in stomatal conductance due to An’ affecting eq. 1? If so, mention this for clarification.

Eq. 4 – I would add a ‘Delta’ symbol to Psi_dehydration, since it seems a water potential drop, rather than a water potential value.

L255 (eq. 5) – I was expecting this equation to relate Psi_dehydration to wdef explicitly, but it does not. Then, wdef accumulates because of gmin and eq. 6a? The way internal water redistribution affects dehydration is not clear either. I would expect PV curves (relating water content to water potential) be used here.

L271 (eq. 6b) – Can these parameters be estimated from standard vulnerability curves?

L304 – Why is Vcmax,25 not mentioned in Table 1? It was calibrated but is not considered a key parameter?

L311 – “BayesianTools”

L360-363 – There are some inconsistencies in this interpretation. If the turning point is behaviour is psi_disconnect, how is it possible that after disconnection stomatal conductance is mostly limited by soil water availability. Then, you state that the dehydration rate depends on gmin and VPD, whereas the evapotranspiration demand was mostly affecting conductance during the wet season. Please, revise these sentences.

L442 – You could be more specific here. Do you refer to acclimation of the pine tree density or leaf area to the climate at Yatir? Or to the general adaptation of P. halepensis as a species, to dry climates?

L476-479 – Note that soil-to-root conductance can strongly decrease but still have your plants connected to the soil. In addition to reduction of conductance (or disconnection, as in your case), one needs explicit (or implicit) water compartments to achieve plant water potentials less negative than soil water potentials in a model, regardless of the complexity of hydraulics.

L496 – Here you could mention other sites (e.g. Puechabon, EucFACE) where litterfall can be more safely attributed to drought and, therefore, would be more amenable for testing the importance of simulating drought-related leaf senescence.

L529 – In SurEau, gmin is dependent on Tleaf (https://gmd.copernicus.org/articles/15/5593/2022/)