Grazing #1140

ckoven · 2023-12-19T16:29:32Z

An initial implementation to allow herbivores to graze and browse plants, with different grazing rates on different land use classes.

Description:

This PR adds the basic capability to have grazers and browsers eat plants. The basic idea is to specify this as a rate of consumption in units of 1/time. I.e. grazers and browsers eat X% of all accessible leaves of edible plants per day on columns of land-use type Y, as controlled by the parameter fates_landuse_grazing_rate(fates_landuseclass). The specification of what is considered edible plants is parametric and controlled by a PFT-level trait fates_landuse_grazing_palatability(fates_pft), so e.g. this can be specified to include grazers (who eat herbaceous plants only) and/or browsers (who eat the leaves of woody plants). And then all leaves of that PFT on columns of that land use type that are below a height of parameter fates_landuse_grazing_maxheight can be subject to herbivory.

The fate of the eaten material is controlled by three "use-efficiency" parameters: fates_landuse_grazing_carbon_use_eff', fates_landuse_grazing_nitrogen_use_eff, 'fates_landuse_grazing_phosphorus_use_eff'. So some mass can be lost from the ecosystem and some can be added to the litter pools via manure (and, I suppose, mortality of grazers).

Default parameter values that are there now (all of which extremely subject to change) are to make this only apply on pasture and rangeland, with a rate of 4% per day based on https://doi.org/10.5194/gmd-11-815-2018 by @samsrabin et al., and currently only grazers (i.e. only applies to nonwoody PFTs). I couldn't find a whole lot of data on the use-efficiency parameters, but those shouldn't be too important until we are running nutrient cycles at large scale under transient change.

fixes #936

Collaborators:

Expectation of Answer Changes:

This will be answer changing if the parameters are set to allow grazing.

Checklist

If this is your first time contributing, please read the CONTRIBUTING document.

All checklist items must be checked to enable merging this pull request:

Contributor

The in-code documentation has been updated with descriptive comments
The documentation has been assessed to determine if updates are necessary

Integrator

FATES PASS/FAIL regression tests were run
Evaluation of test results for answer changes was performed and results provided

Documentation

Technical Note update:
User's Guide update:

Test Results:

CTSM (or) E3SM (specify which) test hash-tag:

CTSM (or) E3SM (specify which) baseline hash-tag:

FATES baseline hash-tag:

Test Output:

samsrabin

Not sure if this was actually ready for review, as it has the "draft" tag but isn't a Draft in the Github sense. Anyway, I've made some (initial?) comments. Also: Where are the new parameters actually read?

samsrabin · 2024-07-05T21:57:54Z

biogeochem/EDPhysiologyMod.F90

+
+       currentSite%mass_balance(element_id)%herbivory_flux_out = &
+            currentSite%mass_balance(element_id)%herbivory_flux_out + &
+            leaf_herbivory * (1._r8 - herbivory_element_use_efficiency) * currentCohort%n


I'm guessing herbivory_flux_out is actually sent to the atmosphere by the HLM, right?

it isn't yet, but will be in a future update.

samsrabin · 2024-07-05T22:08:30Z

parteh/PRTGenericMod.F90

+
+    function GetHerbivory(this, organ_id, element_id, position_id) result(herbivory_val)
+
+      ! This function is very very similar to GetBurned


In fact it looks identical, aside from herbivory replacing burned. Could these be replaced with a single function? Named, e.g., GetDisturbed.

samsrabin · 2024-07-05T22:10:48Z

parteh/PRTLossFluxesMod.F90

+          do i_pos = 1,prt_global%state_descriptor(i_var)%num_pos
+
+             ! The mass that is leaving the plant
+             consumed_mass = mass_fraction * prt%variables(i_var)%val(i_pos)


consumed_mass should presumably be herbivore_consumed_mass?

ckoven · 2024-07-08T21:29:08Z

Not sure if this was actually ready for review, as it has the "draft" tag but isn't a Draft in the Github sense. Anyway, I've made some (initial?) comments. Also: Where are the new parameters actually read?

Thanks @samsrabin! Yes, this is draft in the sense that it wasn't done or tested. And yes, one of the things that isn't done yet is the code to actually read the new parameters. But better to get eyes on it sooner rather than later.

rgknox · 2024-11-18T18:23:28Z

We need to bump this value up to something bigger, we are hitting our limit on maximum parameters already:

https://github.com/NGEET/fates/blob/main/main/FatesParametersInterface.F90#L12

Maybe 400?

glemieux

I had some minor suggested name changes and one question. I'll push an update to add the xml patch file for the cdl update.

biogeochem/FatesLandUseChangeMod.F90

main/EDMainMod.F90

main/FatesHistoryInterfaceMod.F90

Bug fix for issue NGEET#1301 Corrects when the fraction of burnt fuel is zero'd and how the litter mass is updated based on that fraction.

glemieux · 2025-01-21T22:26:48Z

Preliminary regression testing against sci.1.80.6_api.37.0.0-ctsm5.3.018 baseline shows expected FIELDLIST differences due to the new FATES_GRAZING history output not being available in the baselines. Almost all other tests are B4B, with the exception of the FatesColdPRT2 test:

  525 dec1212.hsn.de.hpc.ucar.edu 622: forrtl: severe (408): fort: (2): Subscript #1 of the array MASS_BALANCE has value 4 which is greater than the upper bound of 3
  526 dec1212.hsn.de.hpc.ucar.edu 622:
  527 dec1212.hsn.de.hpc.ucar.edu 622: Image              PC                Routine            Line        Source
  528 dec1212.hsn.de.hpc.ucar.edu 622: cesm.exe           0000000001D15A0B  edphysiologymod_m        2964  EDPhysiologyMod.F90
  529 dec1212.hsn.de.hpc.ucar.edu 622: cesm.exe           0000000001CD062B  edphysiologymod_m         485  EDPhysiologyMod.F90
  530 dec1212.hsn.de.hpc.ucar.edu 622: cesm.exe           0000000001BFE5BB  edmainmod_mp_ed_i         785  EDMainMod.F90
  531 dec1212.hsn.de.hpc.ucar.edu 622: cesm.exe           0000000001BF7DB2  edmainmod_mp_ed_e         223  EDMainMod.F90
  532 dec1212.hsn.de.hpc.ucar.edu 622: cesm.exe           0000000000B13944  clmfatesinterface        1250  clmfates_interfaceMod.F90
  533 dec1212.hsn.de.hpc.ucar.edu 622: cesm.exe           0000000000A89E08  clm_driver_mp_clm        1143  clm_driver.F90

I haven't dug into this yet. Any initial thoughts @rgknox?

Results: /glade/u/home/glemieux/scratch/ctsm-tests/tests_pr1140-fates-woodyfix

ckoven · 2025-01-22T00:10:33Z

Preliminary regression testing against sci.1.80.6_api.37.0.0-ctsm5.3.018 baseline shows expected FIELDLIST differences due to the new FATES_GRAZING history output not being available in the baselines. Almost all other tests are B4B, with the exception of the FatesColdPRT2 test:

  525 dec1212.hsn.de.hpc.ucar.edu 622: forrtl: severe (408): fort: (2): Subscript #1 of the array MASS_BALANCE has value 4 which is greater than the upper bound of 3
  526 dec1212.hsn.de.hpc.ucar.edu 622:
  527 dec1212.hsn.de.hpc.ucar.edu 622: Image              PC                Routine            Line        Source
  528 dec1212.hsn.de.hpc.ucar.edu 622: cesm.exe           0000000001D15A0B  edphysiologymod_m        2964  EDPhysiologyMod.F90
  529 dec1212.hsn.de.hpc.ucar.edu 622: cesm.exe           0000000001CD062B  edphysiologymod_m         485  EDPhysiologyMod.F90
  530 dec1212.hsn.de.hpc.ucar.edu 622: cesm.exe           0000000001BFE5BB  edmainmod_mp_ed_i         785  EDMainMod.F90
  531 dec1212.hsn.de.hpc.ucar.edu 622: cesm.exe           0000000001BF7DB2  edmainmod_mp_ed_e         223  EDMainMod.F90
  532 dec1212.hsn.de.hpc.ucar.edu 622: cesm.exe           0000000000B13944  clmfatesinterface        1250  clmfates_interfaceMod.F90
  533 dec1212.hsn.de.hpc.ucar.edu 622: cesm.exe           0000000000A89E08  clm_driver_mp_clm        1143  clm_driver.F90

I haven't dug into this yet. Any initial thoughts @rgknox?

Results: /glade/u/home/glemieux/scratch/ctsm-tests/tests_pr1140-fates-woodyfix

I think I see the error that caused this, will fix.

biogeochem/FatesLandUseChangeMod.F90

main/EDMainMod.F90

CamelCase! (camels are grazers too) Co-authored-by: Gregory Lemieux <[email protected]>

jenniferholm

Thanks Charlie this is really cool work. My biggest question was just some confusion around if mass can be loss from the ecosystem via both food/consumption pathway and to the litter flux to the atmosphere, and added to the litter via manure. Are all three pathways occurring, and is the order correct? Plus does litter from natural litter fluxes plus manure equal the total litter in the history outputs? I'm not sure that made sense, but wondering if we could separate those out? Thanks!

jenniferholm · 2025-01-25T03:13:19Z

biogeochem/EDPhysiologyMod.F90

@@ -2849,6 +2851,8 @@ subroutine CWDInput( currentSite, currentPatch, litt, bc_in)
    integer  :: numlevsoil        ! Actual number of soil layers

    real(r8) :: SF_val_CWD_frac_adj(4) !SF_val_CWD_frac adjusted based on cohort dbh
+    real(r8) :: leaf_herbivory
+    real(r8) :: herbivory_element_use_efficiency
    !----------------------------------------------------------------------


Maybe give a short definition of these two? leaf_herbivory and herbivory_element_use_efficiency

jenniferholm · 2025-01-25T03:29:25Z

biogeochem/EDPhysiologyMod.F90

@@ -2928,7 +2943,9 @@ subroutine CWDInput( currentSite, currentPatch, litt, bc_in)
       do dcmpy=1,ndcmpy
          dcmpy_frac = GetDecompyFrac(pft,leaf_organ,dcmpy)
          litt%leaf_fines_in(dcmpy) = litt%leaf_fines_in(dcmpy) + &
-               (leaf_m_turnover+repro_m_turnover) * plant_dens * dcmpy_frac
+               (leaf_m_turnover+repro_m_turnover + &
+               leaf_herbivory * herbivory_element_use_efficiency) * &


Just to help me understand the steps, will the leaf litter here already take into account the leaf mass that is consumed and "stays" in the animal as food, and only a portion returning to the litter as "manure"? Is that the herbivory_element_use_efficiency parameter, or something else? Because I'm also confused about what portion of the litter is food, is manure, and/or to the atmosphere, and if all 3 paths are tracked?

The actual grazing animals here (as with all heterotrophs in the soil BGC code) aren't tracked, they only exist as a transformation of leaf biomass to either litter (proportional to herbivory_element_use_efficiency) or the atmosphere (proportional to 1- herbivory_element_use_efficiency) . So this pathway here is the fraction that is returned to litter. In principle this is via manure (and I guess some fraction of animals die while grazing), but for simplicity we are just tracking it as flux into the leaf litter pool, so that it gets a shallow vertical profile like other sources of leaf litter.

jenniferholm · 2025-01-25T03:30:37Z

biogeochem/EDPhysiologyMod.F90

+
+       site_mass%herbivory_flux_out = &
+            site_mass%herbivory_flux_out + &
+            leaf_herbivory * (1._r8 - herbivory_element_use_efficiency) * currentCohort%n


Since it looks like 1._r8 - herbivory_element_use_efficiency is the portion that goes to the atmosphere, is the other portion only going to manure, food diet consumption, or both? My bad if I'm not following

no food diet here. I guess all trophic levels above the grazers (i.e. including humans) should be considered lumped into the herbivory_element_use_efficiency term.

jenniferholm · 2025-01-25T03:36:46Z

main/FatesHistoryInterfaceMod.F90

@@ -6764,6 +6770,12 @@ subroutine define_history_vars(this, initialize_variables)
            upfreq=group_dyna_simple, ivar=ivar, initialize=initialize_variables,                &
            index = ih_npp_stor_si)

+       call this%set_history_var(vname='FATES_GRAZING', units='kg m-2 s-1',    &
+            long='grazing by herbivores of leaves in kg carbon per m2 per second',          &


This is great to have this FATES_GRAZING output and know the mass of leaves eaten. In the future could we also have a FATES_GRAZING_LITTER_OUT type output, that gives us an idea of the amount of manure output. Or helps to differentiate how much litter is natural from turnover vs. litter from herbivores. And maybe useful later for CH4 tracking?

Yes good idea. Will add to to-do list.

jenniferholm · 2025-01-25T03:38:20Z

parameter_files/fates_params_default.cdl

+
+ fates_landuse_grazing_nitrogen_use_eff = 0.25 ;
+
+ fates_landuse_grazing_phosphorus_use_eff = 0.5 ;


Are these _use_eff values just placeholders, or from literature (Sam's paper)? Why again is the carbon_use_eff = 0 ?

These values are fairly arbitrary right now. I am only testing the simplest value, which is carbon-only and where all grazed carbon is lost to the atmosphere. Will parametrically test the consequences of adding these other pathways and nutrients in time.

jenniferholm · 2025-01-25T03:41:39Z

parteh/PRTLossFluxesMod.F90

+    ! with herbivory. There is only one destiny for eaten mass within
+    ! the organ, and that is outside the plant.  
+    ! It is also assumed that non PARTEH parts of the code (ie the grazing-model)
+    ! will decide what to do with the consumed mass (i.e. sent it to the litter


Just for clarity and ease of understanding, maybe explicitly say where the “grazing-model” is for tracking and clarity. Is this “subroutine FatesGrazing” you mean or something else?

ckoven

Thanks @jenniferholm for these questions. Tried to update the documentation to capture these points.

biogeochem/EDPhysiologyMod.F90

ckoven · 2025-01-27T19:44:59Z

biogeochem/EDPhysiologyMod.F90

@@ -2928,7 +2943,9 @@ subroutine CWDInput( currentSite, currentPatch, litt, bc_in)
       do dcmpy=1,ndcmpy
          dcmpy_frac = GetDecompyFrac(pft,leaf_organ,dcmpy)
          litt%leaf_fines_in(dcmpy) = litt%leaf_fines_in(dcmpy) + &
-               (leaf_m_turnover+repro_m_turnover) * plant_dens * dcmpy_frac
+               (leaf_m_turnover+repro_m_turnover + &
+               leaf_herbivory * herbivory_element_use_efficiency) * &


The actual grazing animals here (as with all heterotrophs in the soil BGC code) aren't tracked, they only exist as a transformation of leaf biomass to either litter (proportional to herbivory_element_use_efficiency) or the atmosphere (proportional to 1- herbivory_element_use_efficiency) . So this pathway here is the fraction that is returned to litter. In principle this is via manure (and I guess some fraction of animals die while grazing), but for simplicity we are just tracking it as flux into the leaf litter pool, so that it gets a shallow vertical profile like other sources of leaf litter.

ckoven · 2025-01-27T19:45:16Z

biogeochem/EDPhysiologyMod.F90

@@ -2849,6 +2851,8 @@ subroutine CWDInput( currentSite, currentPatch, litt, bc_in)
    integer  :: numlevsoil        ! Actual number of soil layers

    real(r8) :: SF_val_CWD_frac_adj(4) !SF_val_CWD_frac adjusted based on cohort dbh
+    real(r8) :: leaf_herbivory
+    real(r8) :: herbivory_element_use_efficiency
    !----------------------------------------------------------------------


ckoven · 2025-01-27T19:45:50Z

biogeochem/EDPhysiologyMod.F90

+
+       currentSite%mass_balance(element_id)%herbivory_flux_out = &
+            currentSite%mass_balance(element_id)%herbivory_flux_out + &
+            leaf_herbivory * (1._r8 - herbivory_element_use_efficiency) * currentCohort%n


it isn't yet, but will be in a future update.

ckoven · 2025-01-27T19:47:15Z

biogeochem/EDPhysiologyMod.F90

+
+       site_mass%herbivory_flux_out = &
+            site_mass%herbivory_flux_out + &
+            leaf_herbivory * (1._r8 - herbivory_element_use_efficiency) * currentCohort%n


no food diet here. I guess all trophic levels above the grazers (i.e. including humans) should be considered lumped into the herbivory_element_use_efficiency term.

ckoven · 2025-01-27T19:48:12Z

main/FatesHistoryInterfaceMod.F90

@@ -6764,6 +6770,12 @@ subroutine define_history_vars(this, initialize_variables)
            upfreq=group_dyna_simple, ivar=ivar, initialize=initialize_variables,                &
            index = ih_npp_stor_si)

+       call this%set_history_var(vname='FATES_GRAZING', units='kg m-2 s-1',    &
+            long='grazing by herbivores of leaves in kg carbon per m2 per second',          &


Yes good idea. Will add to to-do list.

ckoven · 2025-01-27T19:49:40Z

parameter_files/fates_params_default.cdl

+
+ fates_landuse_grazing_nitrogen_use_eff = 0.25 ;
+
+ fates_landuse_grazing_phosphorus_use_eff = 0.5 ;


These values are fairly arbitrary right now. I am only testing the simplest value, which is carbon-only and where all grazed carbon is lost to the atmosphere. Will parametrically test the consequences of adding these other pathways and nutrients in time.

ckoven · 2025-01-27T19:51:13Z

parteh/PRTLossFluxesMod.F90

+    ! with herbivory. There is only one destiny for eaten mass within
+    ! the organ, and that is outside the plant.  
+    ! It is also assumed that non PARTEH parts of the code (ie the grazing-model)
+    ! will decide what to do with the consumed mass (i.e. sent it to the litter


parteh/PRTLossFluxesMod.F90

glemieux · 2025-01-30T00:50:35Z

Preliminary regression testing against sci.1.80.6_api.37.0.0-ctsm5.3.018 baseline shows expected FIELDLIST differences due to the new FATES_GRAZING history output not being available in the baselines. Almost all other tests are B4B, with the exception of the FatesColdPRT2 test:

Rerunning the full fates regression tests results in all passing now. Note that these were not tested against a baseline as these are intended as preliminary.

ckoven and others added 13 commits May 1, 2023 21:08

adding land use parameters to file: grazing, PFT max canopy coverage

ee5ef88

Merge branch 'luh2' into grazing

bc633e2

starting to add grazing parameterization

dbd692d

more grazing updates

580df9e

Merge branch 'luh2' into grazing_merge

c62e98d

merged luh2 and resolved conflicts

cab4ee7

fixed merged conflict

cb5ea6d

Merge branch 'luh2' into grazing_merge

dcbf787

Merge branch 'luh2' into grazing_merge

2b36e09

Merge branch 'luh2' into grazing_merge

9e035dd

merged main and resolved conflicts

3f755f0

changed hite to height

4e818e8

changed grazing rate to 4%/day

b67019d

ckoven added the draft label Dec 19, 2023

glemieux requested a review from samsrabin May 28, 2024 18:44

glemieux requested a review from jenniferholm June 17, 2024 19:24

samsrabin reviewed Jul 5, 2024

View reviewed changes

ckoven added 2 commits July 19, 2024 11:57

merged main and resolved conflicts

53a6a44

read the grazing parameters

636e003

glemieux added the parameter file Pertaining to changes to the FATES parameter file label Jul 22, 2024

ckoven added 3 commits July 24, 2024 09:37

set all grazing_rate parameters to zero

2b87e44

compile-time bugfixes

c039964

added grazing carbon flux history variable

e7b236b

ckoven removed the draft label Oct 31, 2024

glemieux requested changes Jan 11, 2025

View reviewed changes

biogeochem/FatesLandUseChangeMod.F90 Outdated Show resolved Hide resolved

main/EDMainMod.F90 Outdated Show resolved Hide resolved

main/FatesHistoryInterfaceMod.F90 Show resolved Hide resolved

add api37 parameter file and xml patch file

ac5bf38

glemieux added 2 commits January 13, 2025 15:27

Merge tag 'sci.1.80.6_api.37.0.0' into grazing_merge

b3113b7

Bug fix for issue NGEET#1301 Corrects when the fraction of burnt fuel is zero'd and how the litter mass is updated based on that fraction.

correct woody logic check

6fa3079

fix error to EDPhysiologyMod.F90 when nutrients are on

4b100b2

ckoven commented Jan 24, 2025

View reviewed changes

biogeochem/FatesLandUseChangeMod.F90 Outdated Show resolved Hide resolved

ckoven commented Jan 24, 2025

View reviewed changes

biogeochem/FatesLandUseChangeMod.F90 Outdated Show resolved Hide resolved

ckoven commented Jan 24, 2025

View reviewed changes

main/EDMainMod.F90 Outdated Show resolved Hide resolved

Apply suggestions from code review

76225c9

CamelCase! (camels are grazers too) Co-authored-by: Gregory Lemieux <[email protected]>

jenniferholm reviewed Jan 25, 2025

View reviewed changes

This was referenced Jan 25, 2025

Add description of the land use grazing mode NGEET/fates-docs#70

Open

Add grazing mode subsection NGEET/fates-users-guide#83

Open

ckoven commented Jan 27, 2025

View reviewed changes

ckoven and others added 2 commits January 27, 2025 12:28

Apply suggestions from code review

b619100

added herbivory flux to restart data

da70f55

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Grazing #1140

Grazing #1140

ckoven commented Dec 19, 2023

samsrabin left a comment

samsrabin Jul 5, 2024

ckoven Jan 27, 2025

samsrabin Jul 5, 2024

samsrabin Jul 5, 2024

ckoven commented Jul 8, 2024

rgknox commented Nov 18, 2024

glemieux left a comment

glemieux commented Jan 21, 2025 •

edited

Loading

ckoven commented Jan 22, 2025

jenniferholm left a comment

jenniferholm Jan 25, 2025

ckoven Jan 27, 2025

jenniferholm Jan 25, 2025

ckoven Jan 27, 2025

jenniferholm Jan 25, 2025

ckoven Jan 27, 2025

jenniferholm Jan 25, 2025

ckoven Jan 27, 2025

jenniferholm Jan 25, 2025

ckoven Jan 27, 2025

jenniferholm Jan 25, 2025

ckoven Jan 27, 2025

ckoven left a comment

ckoven Jan 27, 2025

ckoven Jan 27, 2025

ckoven Jan 27, 2025

ckoven Jan 27, 2025

ckoven Jan 27, 2025

ckoven Jan 27, 2025

ckoven Jan 27, 2025

glemieux commented Jan 30, 2025


		function GetHerbivory(this, organ_id, element_id, position_id) result(herbivory_val)

		! This function is very very similar to GetBurned


		fates_landuse_grazing_nitrogen_use_eff = 0.25 ;

		fates_landuse_grazing_phosphorus_use_eff = 0.5 ;

Grazing #1140

Are you sure you want to change the base?

Grazing #1140

Conversation

ckoven commented Dec 19, 2023

Description:

Collaborators:

Expectation of Answer Changes:

Checklist

Documentation

Test Results:

samsrabin left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

ckoven commented Jul 8, 2024

rgknox commented Nov 18, 2024

glemieux left a comment

Choose a reason for hiding this comment

glemieux commented Jan 21, 2025 • edited Loading

ckoven commented Jan 22, 2025

jenniferholm left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

ckoven left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

glemieux commented Jan 30, 2025

glemieux commented Jan 21, 2025 •

edited

Loading