Has the solute node (or salinity-related modules/processes) been activated/applied in the APSIM-Lucerne model currently?

[Submarine-666]

Hi APSIM community,

The version we're now using is Classic 7.10. In order to simulate the growth of alfalfa under soil salinity stress, I inserted "solute" node into the simulation tree, and also input initial EC values in the "analysis" node under "soil" node. However, we have no idea if this solute node (or salinity-related modules/processes) is available/exercisable or been activated in the APSIM-Lucerne model?

If yes, we guess the salt stress is regarded as an additional water stress, by imposing a osmotic stress on some growth processes, e.g., water uptake. Our questions are: which relevant processes are affected by this osmotic stress? where could we found the relationship between the "osmotic stress" and "EC/Cloride"?

Thanks a lot!

[peter-devoil]

The simplest way to find out is to set the solute levels to obscene levels and see if the simulated plant still grows - I suspect it will pay no attention to it.

The wheat module has code to reduce kl - the rate of water uptake - in response to EC/Cl; a species specific relation. I think it's described here: https://doi.org/10.1071/AR06365 and S11.6 of https://www.apsim.info/wp-content/uploads/2019/09/WheatDocumentation.pdf

Likewise ll - the lower limit of extraction - is a candidate for modification.

You could modify these parameters "by hand" for your soil but remain aware that these modifications are hard to defend without supporting / experimental data.

[Submarine-666]

We totally agree with your idea and we found that the kl_modifier has already been realized but only in the Wheat model. And it seemed like soil salinity had not effect on alfalfa growth in the Lucerne model currently. We guess we could only modify these parameters (e.g. CLL, KL) "by hand" at the very beginning on the user interface, which is not a time-dynamic method of mimicking alfalfa growth. We realized it could be challenging to handle the code things, but we're wondering: has there any other parameter been modified under stress (e.g. RUE, TE) in other crop models of APSIM? Cheers.

[peter-devoil]

The "philosophy" behind the model is to avoid modifying variables unless there is justification: if your plants have very different genetics (eg landrace vs hybrids) then RUE/TE might be candidates for modification, but if they're the same genetics then its hard to consider they should be different (especially if its due to some soil constraint). Again; an experiment is the definative answer, not field observations or guesses.

The EC/Cl modifications of ll/kl are directly related to the plant's environment, don't change through the lifetime of the plant, and are only tedious by their number.

For some encouragement: I can share that a number of our ML experiments in preparation are saying that we can detect differences in canopy development due to total water holding capacity, but not any difference in water distribution. This might be a "counting angels on a pin" problem..

[Submarine-666]

Thanks for your reply. We've conducted a three-year field experiment, aiming to depict alfalfa growth under different water stress. Different treatments of irrigation amount and soil salinity were set with only one cultivar. We found there's some overestimation of biomass and underestimation of soil water content under severe water and salt stress, and after searching relevant literatures and conducting sensitivity analysis, we suppose the LL/KL, RUE, and TE might be the critical factors that have been affected by the stress.

By the "EC/Cl modifications of ll/kl" part you mentioned, do you mean the kl_modification that had been realized in Wheat model is not time-dynamic (kl values are state variables along the lifetime under different salt stress) currently?

We'll appreciate it if you could give us more comments about differences in canopy development you mentioned and other things. You could contact me at yanzhe.hu@slu.se

Cheers

[BrianCollinss]

If you are interested in calibrating APSIM for your experiment, I will be happy to run the calibration for you. I have developed an algorithm to do that in R (https://onlinelibrary.wiley.com/doi/abs/10.1111/jac.12575). In case you are interested in this collaboration, contact me on brian.collins@jcu.edu.au.

[peter-devoil]

Hi Yanzhe - the kl modification is time dynamic (calculated every day) - and in some environments, eg saline water tables, could change from day to day. I apologise for thinking only of the environment at my door..

You may want to explore the kl modifier - it's an optional variable in the .ini file (there's a brief description of how to edit it here). Search the wheat.xml file for "ModifyKL, ClA, ClB ..." and add the same to lucerne.xml.

Again, test the model to be sure your changes are having an effect.

[BrianCollinss]

The paper to read is:
Hochman et. al. (2007) Simulating the effects of saline and sodic subsoils on wheat crops growing on Vertosols. Australian Journal of Agricultural Research, 58, 802:810

[Submarine-666]

@BrianCollinss Thanks for your comments. I read your article about the optimization method used to calibrate parameters of APSIM, which inspired us a lot. We also referred to the article (Hochman et al., 2007), and are managing to realize the modification under salt stress for alfalfa. We'll contact you if we run into trouble again. Thanks.

[Submarine-666]

Hi experts,

We are now managing to modify kl under stress, by establishing a relationship between klfactor and Cl, but we are confused that the unit of Cl (mg/kg), does it mean mg Cl per kg water, or mg Cl per kg soil?

Thanks.