Climate change will affect both the mean and variability in environmental conditions and may have major, negative impacts on population densities in the future. For annual plants that already live in an extreme environment like the Sonoran Desert, keeping a fraction of their seeds dormant underground (for possibly years at a time) is critical to survive. Here, we consider how this form of bet-hedging (i.e., delayed germination) for ten Sonoran Desert annuals mediates responses to precipitation shifts. We use a demographic model parameterized with long-term field and precipitation data to explore how forecasted changes in precipitation impact annual plant species’ population densities. We then examine how instantaneous adaptation to optimal germination fractions in the shifted precipitation regimes bolsters population densities. Our results indicate that overall less rainfall and, to a lesser extent, increased variance in rainfall drive population levels down. Instantaneous adaptation to optimal germination fractions in new regimes only benefited species' populations marginally and only for small-to-moderate shifts in precipitation. Thus, even rapid adaptation is unlikely to save populations experiencing larger shifts in precipitation. Finally, species exhibiting both higher dormant seed survivorships and abilities to capitalize on wet-year bonanzas were the most resilient to precipitation shifts and benefited relatively the most from instantaneous adaptation to optimal germination fractions.

---

CORRESPONDING AUTHOR INFORMATION

NAME: William S. Cuello
CORRESPONDING AUTHOR EMAIL (Los Alamos National Laboratory email; current institution of employment): wscuello@lanl.gov,
ALTERNATE EMAIL (if not reachable via above email): wscuello@gmail.com

---

R INFO

VERSION: 2022.07.1+554
PACKAGES: leaps, viridis, gridExtra, ggplot2

---

ORDERING OF THE FOLLOWING SECTIONS

SECTION 1: R SCRIPT CALL FILES SECTION
SECTION 2: R SCRIPT FIGURE FILES SECTION
SECTION 3: CSV FILES

Work Flow:

If starting from scratch:

1. Run master_flag_caller_HIREC_mean_only.R to have specified number of runs (see within script: variable name -- num_of_source_retrials; type -- integer). This will generate a data file similar to "mean_only flag= 1 init_run_size= 5000 time_forward= 5000 N_amount= 2". Here, variables init_run_size and time_forward can be changed in master_flag_caller_HIREC_mean_only.R. This file allows you to call figures_HIREC_v2.R, which generates Figures 1-4 and Appendices B and F.
2. Obtaining all other files for Figures 5-7 and Appendices A and C require modification of master_flag_caller_HIREC_mean_only.R. In particular, to perform the sensitivity analyses, one has to go to use_table within pre_processing_HIREC_mean_only.R and manually subtract or add 10% to the point-value listed under the correct category:

survs -> dormant seed survivorships survs_new -> fresh seed
survivorships a_comp -> intraspecific competition alphas -> dry-year
seed yield betas -> sensitivity of seed yield to precipitation LH_b0s
-> dry-year chance of reproduction LH_b1s -> sensitivity of chance of
reproduction to precipitation

You will have to change the save.image(...) file name or change the file
directory, so that it does not overwrite your previous save. With these
files saved, you can now run
HIREC_amnat_post_processing_and_figures_positive.R for all 10% increases
in life-history traits and functional responses to precipitation; you
can run HIREC_amnat_post_processing_and_figures_negative.R for
reductions in these parameters. These two files are similar but require
both data files generated for increases in standard deviation in
precipitation and mean reductions in precipitation. In particular, one
will have to repeat process 1), 2) and onward with
master_flag_caller_HIREC_SD_only.R.

Then to generate the rest of the figures (in Appendix C) one
must run quasi_extinction_probabilities_experiment_QEP_mean.R for
quasi-extinction probability pdfs;
quasi_extinction_probabilities_experiment_PTB_type_1_mean.R for percent
time below the historical threshold pdfs; and
quasi_extinction_probabilities_experiment_PTB_type_2_mean_scenario.R
for percent time below a fraction of the original density pdfs.

NOTE: CREATION OF THE DATA (USING CLUSTERS) IS CURRENTLY ONLY COMPATIBLE
WITH UBUNTU AND LINUX-BASED SYSTEMS.
_______________________________________________________________________________

If starting from files available:

One can already generate figures from figures_HIREC_v2.R but will still
need to generate the Appendix figures in Appendix C using
quasi_extinction_probabilities_experiment_QEP_mean.R,
quasi_extinction_probabilities_experiment_PTB_type_1_mean.R, and
quasi_extinction_probabilities_experiment_PTB_type_2_mean_scenario.R.

%%%%%%%%%%%%%%%%%%%%%%% % R SCRIPT CALL FILES % %%%%%%%%%%%%%%%%%%%%%%%

File name: post_processing_HIREC.R

Description: creates all the statistics for percent time below the
historical threshold, ESS germination fraction, and reduction in PTB
plots that are generated in the script figures_HIREC_v2.R.
_______________________________________________________________________________

File name: 18.10.21_paper1_calledon.R

Description: All functions that are needed to iterate a population (or
populations in the case of multiple lineages) forward (i.e., Eq.(1)).
ESS_func and mean_drdg2 are used to obtain the ESS value (i.e., Eq.(6)
and the root of the expression).
_______________________________________________________________________________

File name: pre_processing_HIREC_SD_only.R

Description: This is the main file for manipulating precipitation
regimes and finding population densities in each precipitation regime,
calculating percentage time below the historical threshold (PTB),
evolutionarily stable strategies for germination rates, and reduction in
PTB due to adaptation. In particular, we grab life-history traits (e.g.,
survivorships and intraspecific competition rates) taken from Gremer
2014. We relate precipitation to yield using the hurdle model (see
manuscript and Cuello 2019). We calculate evolutionarily stable
strategies in from the lab-measured precipitation values (see manuscript
and Cuello 2019). We run each species' population forward, drawing from
a log-normal distribution fitted to the precipitation values (which we
call historical precipitation regime). We calculate the per-species
historical threshold from the distribution of density values generated
under the historical precipitation regime. In this file, we ONLY
manipulate precipitation's variance in this file and calculate the PTB,
ESS values, and reduction in PTB in each regime.

WARNING: Does not work in WINDOWS -- need MAC OS.

---

File name: pre_processing_HIREC_mean_only.R

Description: The first part of this file is exactly the same as
pre_processing_HIREC_SD_only.R. The only difference between these two
files: we ONLY reduce mean precipitation in this file and calculate the
PTB, ESS values, and reduction in PTB in each regime.

WARNING: Does not work in WINDOWS -- need MAC OS.
_______________________________________________________________________________

File name: master_flag_caller_HIREC_amnat_SD_5k.R

Description: Calls pre_processing_HIREC_SD_only.R. This file can be
called once or multiple times, depending on the context. For example, if
one wants to average over multiple runs, the matrices are quite large;
therefore, one needs to keep calling pre_processing_HIREC_SD_only.R in
order to get multiple runs (for independent draws of precipitation ->
different draws of seed yield per species -> different PTBs if the
length of the run is small enough (e.g., Tf = 50)).

WARNING: Does not work in WINDOWS -- need MAC OS.
_______________________________________________________________________________

File name: master_flag_caller_HIREC_amnat_mean_5k.R

Description: Calls pre_processing_HIREC_mean_only_5k.R. This file can be
called once or multiple times, depending on the context. For example, if
one wants to average over multiple runs, the matrices are quite large;
therefore, one needs to keep calling pre_processing_HIREC_mean_only.R in
order to get multiple runs (for independent draws of precipitation ->
different draws of seed yield per species -> different PTBs if the
length of the run is small enough (e.g., Tf = 50)).

WARNING: Does not work in WINDOWS -- need MAC OS.

%%%%%%%%%%%%%%%%%%%%%%%%% % R SCRIPT FIGURE FILES %%%%%%%%%%%%%%%%%%%%%%%%%%

File name: HIREC_amnat_post_processing_and_figures_negative.R

Figures generated: Appendix A1-A3; A7-A9 Figures

Description: In short, this file generates all the sensitivity analyses that are
performed on the percent time below the historical threshold (PTB) curves, the
evolutionarily stable strategy (ESS) for germination fraction curves,
and reduction in PTB curves when seed survivorships, sensitivity of
chance of reproduction to precipitation, and sensitivity of seed yield
to precipitation are reduced.

---

File name: HIREC_amnat_post_processing_and_figures_positive.R

Figures generated: Figure 4-6; Appendix A4-A6

Description: In short, this file generates all the sensitivity analyses that
are performed on the percent time below the historical threshold (PTB) curves,
the evolutionarily stable strategy (ESS) for germination fraction curves,
and reduction in PTB curves when life-history traits and functional
responses to precipitation are increased.

_______________________________________________________________________________

File name: figures_HIREC_v2.R

Description: Quick visualization of Figures 1-4 in the main text. Also for visualizing
Appendix B and Appendix D.

---

File name: HIREC_appendixCstar

Description: Quick visualization of Appendix C Figures.

---

File name: quasi_extinction_probabilities_experiment_PTB_type_1_SD.R

Description: Generates Appendix C Figure C1B/2B. Run after generating all
sensitivity files by modifying master_flag_caller_HIREC_amnat_SD.R (and
master_flag_caller_HIREC_amnat_SD.R).

---

File name: quasi_extinction_probabilities_experiment_PTB_type_1_mean.R

Description: Generates Appendix C Figure C1A/2A. Run after generating all
sensitivity files by modifying master_flag_caller_HIREC_amnat_mean.R
(and master_flag_caller_HIREC_amnat_mean.R).

---

File name: quasi_extinction_probabilities_experiment_PTB_type_2_SD.R

Description: Generates Appendix C1D/2D. Run after generating
all sensitivity files by modifying master_flag_caller_HIREC_amnat_SD.R
(and master_flag_caller_HIREC_amnat_SD.R).

---

File name: quasi_extinction_probabilities_experiment_PTB_type_2_mean.R

Description: Generates Appendix C1C/2C. Run after generating
all sensitivity files by modifying master_flag_caller_HIREC_amnat_mean.R
(and master_flag_caller_HIREC_amnat_mean.R).

---

File name: quasi_extinction_probabilities_experiment_QEP_SD.R

Description: Generates Appendix C1F/2F. Run after generating
all sensitivity files by modifying master_flag_caller_HIREC_amnat_SD.R
(and master_flag_caller_HIREC_amnat_SD.R).

---

File name: quasi_extinction_probabilities_experiment_QEP_mean.R

Description: Generates Appendix C1E/2E. Run after generating
all sensitivity files by modifying master_flag_caller_HIREC_amnat_mean.R
(and master_flag_caller_HIREC_amnat_mean.R).

%%%%%%%%%%%%%%% %% CSV FILES %% %%%%%%%%%%%%%%%

FILE NAME: DataFig2Venable2007.csv

Column Name: Year Description: Year of the experiment

Column Name: Jmprec Total winter precipitation (mm) -- used in the
script.

Column Name: Fprec Non-winter precipitation years (mm) -- not used in
the script.

Column Name: ln(lxbx+0.5) Per-capita yield of a species' population
being tracked.

Column Name: SP Name of the species name (see Table 1 of manuscript to
see all species names).

---

FILE NAME: LTERDATA1982_2016.csv

Date: Month/Day/Year Date of the collection. ########### refer to no
collection.

Column Name: Precip (total mm) Total precipitation collected that day at
the Tumamoc Hill Laboratory.

Site/Station (CAP/1): (NOT USED) Location where the precipitation values
(mm) were taken from.

Flag (G): (NOT USED) Location of measurement taken (ground
measurements).

---

FILE NAME: SpeciesByYear.csv

Column Names:

1. species
2. year
3. germ
4. survive
5. seedlings_per_m2
6. lx
7. bx
8. lxbx
9. seeds_per_m2
10. germ fraction

Descriptions: This file is taken directly from:
http://www.eebweb.arizona.edu/faculty/venable/LTREB/LTREB%20data.htm#data
Every column name is detailed at this link:
http://www.eebweb.arizona.edu/faculty/venable/LTREB/metadata.htm.

Note: the "." entries is synonymous to leaving the cell blank (i.e. the data is not available). These columns are unnecessary in the sense that they are not used in the models or simulations; however, the R scripts account for these two columns, so they must be kept and** not deleted** upon importing.

---

FILE NAME: threshold_baseline_txt.csv

Column name: threshold_values

Description: the lowest decile of the species' population densities on a
logarithmic scale.