NECN Maximum Drought

[erin.c...@gmail.com]

Hello LANDIS-II Users!

As Paul Henne wrote in his November 12, 2015 post, I am compiling species parameters for the NECN (formerly Century) Succession Extension, and am uncertain how to assign Maximum Drought values.  Here is the description from the User Guide:

2.38.6 MaxDrought (double)
If available water falls below zero for a percent of the growing season greater than this value, a species cannot establish. Units: fraction of the growing season (0.0 – 1.0). Lower values indicate species whose establishment is more sensitive to drought.

How to determine this fraction was unclear to me, as it was to Paul in 2015.  Rob Scheller responded to Paul's email stating:

A drought day is one in which available water is less than the wilting point.  The fraction of days is based on the growing season, i.e., percentage of the growing season with available soil water less than the wilting point.

I'm not sure how one would figure out, for a species across its range, what fraction of the time the species has available water less than the wilting point.  As far as I know, the permanent wilting point is usually measured in the lab using pressure plate apparatus when the suction is at 15 bars.  Is there a way to remotely sense this information?  Liang et al. GCB 2017 state:

A drought day is one in which soil moisture is below a critical soil moisture threshold (i.e., the soil wilting point, Pastor & Post, 1986).  Following the algorithms outlined in LANDIS-II code (www.landis-ii.org/developers), we generated a map depicting the fraction of drought days over the landscape using soil texture data and baseline climate data.  

I don't find any specific algorithms at that website, nor do I have a clear picture of how to generate maps depicting drought days based on this description. 

The clearest description I have found of how to calculate drought tolerance is from Henne et al. 2020 which states:

We estimated this value by first calculating mean AET and PET at 4-km resolution for the conterminous United States for the period 1980–2015 (PRISM Climate Group, 2017; see water balance calculations above). We then assigned the drought stress limit as the proportion of months where PET > AET, standardized by the number of months where PET > 0 for the range of each species (Little Jr., 1971).

The above definition seems sensible, but does not specifically mention the wilting point. 

Further, the drought tolerances listed in a handful of papers vary considerably.  Taking P ponderosa as an example: Liang et al. GCB 2017 report 0.43.  Henne et al. J. of Ecology, 2020 report 0.54. Sierra-Diaz et al Scientific Reports 2018 report 0.9.  Spending 90% of the time at the permanent wilting point seems rather high, but that higher value is consistent with the max drought values for co-occurring species given in the Lake Tahoe 2017 github repository (https://github.com/LANDIS-II-Foundation/Project-Lake-Tahoe-2017).

So I have two questions: 

Does the exact value of drought tolerance matter as much as the relative value compared to the other species in the model?  Assuming it is a relative comparison, I could move forward with calculating drought tolerance with, for example, available maps of CWD ( = PET - AET).

Instead, if there is a specific drought tolerance calculation that needs to correspond to LANDIS's internal NECN algorithms, could someone please provide a more detailed description of how the max drought values are calculated?

Thank you in advance!
Erin Conlisk

[Robert Scheller]

Erin,

Apologies for the slow response.

There's a lot here to unpack.  

First, the establishment code is here:  https://github.com/LANDIS-II-Foundation/Extension-NECN-Succession/blob/master/src/Establishment.cs

The code for 'dry days' (mentioned by Liang et al. GCB 2017) is at L516: https://github.com/LANDIS-II-Foundation/Extension-NECN-Succession/blob/master/src/SoilWater.cs

In general, the 'dry days' code was derived from LINKAGES and many of the original parameters also:  https://link.springer.com/content/pdf/10.1007/BF00224636.pdf

See also: https://daac.ornl.gov/MODELS/guides/LINKAGES.html

To answer your specific question:  The exact values are important, more so than the relative values.  As a rule-of-thumb, the probability of establishment works best if the average across decades is < 0.5.  (This also depends on the succession timestep.)  If average Pest is much higher, a species will tend to get over-established (provided there is enough light).

Cheers,

Rob