| Title: | Simulation of Weed Population Dynamics |
|---|---|
| Description: | An environment to simulate the development of annual plant populations with regard to population dynamics and genetics, especially herbicide resistance. It combines genetics on the individual level (Renton et al. 2011) with a stochastic development on the population level (Daedlow, 2015). Renton, M, Diggle, A, Manalil, S and Powles, S (2011) <doi:10.1016/j.jtbi.2011.05.010> Daedlow, Daniel (2015, doctoral dissertation: University of Rostock, Faculty of Agriculture and Environmental Sciences.) |
| Authors: | Christoph v. Redwitz [aut, cre], Friederike de Mol [aut] |
| Maintainer: | Christoph v. Redwitz <[email protected]> |
| License: | GPL-3 |
| Version: | 0.3.3 |
| Built: | 2024-12-04 07:01:28 UTC |
| Source: | CRAN |
link{struc_preparation}
Before calling struc_preparation it is necessary to bring the genetics into the correct form, which is mainly the correct dimensions.
gen_check(Rmx, af, dom, epis)gen_check(Rmx, af, dom, epis)
Rmx |
maximum resistance value, if all gene loci under consideration are homozygous resistant. |
af |
initial frequency of resistance alleles in the population. |
dom |
dominance of resistance alleles. |
epis |
epistasis value, describing the interaction between resistance alleles. |
If no genetics are required af has to be set to NA. The value of dom is adjusted to the length of af. If there are mismatches, the dom is cut or the first given value is repeated to fit the number of given alleles in af. The variables af and dom are corrected and n_loci is created, which is 0 when no genetics is included. If no value is given to epis it is set to 0.
gen_diploid recombines the alleles of weeds with diploid genome.
gen_diploid( start, start_comb = NA, result, newseeds = get0("newseeds", envir = parent.frame(n = 1)), max_vec_length = 1e+07 )gen_diploid( start, start_comb = NA, result, newseeds = get0("newseeds", envir = parent.frame(n = 1)), max_vec_length = 1e+07 )
start |
column names of parental cohorts. |
start_comb |
all named columns in |
result |
name of the results column. |
newseeds |
number of new seeds calculated by |
max_vec_length |
used internally, a technical term, defining the maximum length of vectors to be used. |
gen_diploid assumes independent allele recombination.
A column with the name of result and, if necessary, a column with the name of start_comb is added to the data.frame dfgenotype.
newseeds are always coerced to a whole number.
## Not run: # generate a 'dfgenotype' data.frame: struc_preparation2(Rmx=10, af=c(0.01, 0.2), epis=0, dom=1) #Distribute 10000 individuals of the starting population across the genotypes. #The two gene loci have initial frequencies of 0.01 and 0.8. gen_freq(af=c(0.01,0.8), n_seeds=10000) # The column "initialSB" represents the parent generation, which recombines and #therefore defines the genetics of the new seeds newseeds <- 10000 gen_diploid(start="initialSB", result="followingSB") # If a second cohort is reprodusing in the same time gen_diploid(start=c("initialSB", "followingSB"), start_comb="two_cohorts_combind", result="followingSB2") rm(dfgenotype, mf, newseeds, xprobab) ## End(Not run)## Not run: # generate a 'dfgenotype' data.frame: struc_preparation2(Rmx=10, af=c(0.01, 0.2), epis=0, dom=1) #Distribute 10000 individuals of the starting population across the genotypes. #The two gene loci have initial frequencies of 0.01 and 0.8. gen_freq(af=c(0.01,0.8), n_seeds=10000) # The column "initialSB" represents the parent generation, which recombines and #therefore defines the genetics of the new seeds newseeds <- 10000 gen_diploid(start="initialSB", result="followingSB") # If a second cohort is reprodusing in the same time gen_diploid(start=c("initialSB", "followingSB"), start_comb="two_cohorts_combind", result="followingSB2") rm(dfgenotype, mf, newseeds, xprobab) ## End(Not run)
gen_freq generates the numbers of genotypes at the beginning of the first simulated year.
gen_freq( af, n_seeds, result = "initialSB", distribution = NA, max_vec_length = 1e+07 )gen_freq( af, n_seeds, result = "initialSB", distribution = NA, max_vec_length = 1e+07 )
af |
initial frequency of resistance alleles in the population. |
n_seeds |
initial number of weed seeds. They will be allocated to the genotypes. |
result |
names of the results columns. |
distribution |
defines the proportion with |
max_vec_length |
used internally, a technical term, defining the maximum length of vectors to be used. |
The start values for a model include the initial frequencies of alleles and the initial seedbank, i.e. the number of seeds in the soil. gen_freq allocates the n_seeds individuals to the different genotypes as provided by dfgenotype, which is created by struc_preparation2(). If af == NULL the function gives all seed to the only "genotype". The distribution can be splitted to multiple cohorts using distribution.
If distribution is "equal", n_seeds are distributed equaly amoung the cohorts defindes with result.
Returns a data.frame containing the genotypes and their frequencies provided by dfgenotype.
# generate a 'dfgenotype' data.frame: struc_preparation2(Rmx=10, af=c(0.01,0.8), epis=0, dom=1) #Distribute 10000 individuals of the starting population across the genotypes. #The two gene loci have initial frequencies of 0.01 and 0.8. gen_freq(af=c(0.01,0.8), n_seeds=10000) rm(dfgenotype, mf, xprobab)# generate a 'dfgenotype' data.frame: struc_preparation2(Rmx=10, af=c(0.01,0.8), epis=0, dom=1) #Distribute 10000 individuals of the starting population across the genotypes. #The two gene loci have initial frequencies of 0.01 and 0.8. gen_freq(af=c(0.01,0.8), n_seeds=10000) rm(dfgenotype, mf, xprobab)
link{sel_herb}
Is called within sel_herb() to check for the existence of the necessary data.
herb_check(put, sdrate, thresh, rate)herb_check(put, sdrate, thresh, rate)
put |
probability of a weed to be untouched by the herbicide. |
sdrate |
variance of the herbicide rate reaching the weed. |
thresh |
threshold herbicide rate to kill weeds without resistance. |
rate |
percentage (%) of the registered herbicide dose. |
Utility function internally used. It's used only by sel_herb, and usually there is no reason to change it. Calculates the surviving number of weeds according to the specific genotype.
intern_herbicide(resist, n_samples, put, rate, sdrate, thresh)intern_herbicide(resist, n_samples, put, rate, sdrate, thresh)
resist |
numeric, value of resistance for the genotype (defined by Renton et al. 2011). Shall be |
n_samples |
integer, the number of weeds with one specific genotype. |
put |
probability of a weed to be untouched by the herbicide. |
rate |
percentage (%) of the registered herbicide dose. |
sdrate |
variance of the herbicide rate reaching the weed. |
thresh |
threshold herbicide rate to kill weeds without resistance. |
intern_herbicide is used in sel_herb. Firstly, it calculates the number of weeds that are untouched by the herbicide by chance (probability=put). In the second step, the herbicide rate that reached an individual is compared with the resistance value (calculated after Renton et al. 2011, sel_resist). If the resistance value is lower than the dose, the weed dies. All surviving weeds are summed up.
The number of weeds surviving the herbicide.
Renton, M.; Diggle, A.; Manalil, S. & Powles, S. (2011): Does cutting herbicide rates threaten the sustainability of weed management in cropping systems? Journal of Theoretical Biology, 283, 14-27.
#How many of 1000 weeds of a genotype with resistance value 5.5 survive a herbicide application #with full dose? 'It is assumed that weeds reseaving less than 20 \% of the full dose survive #independently of their resistant value. intern_herbicide(resist=5.5, n_samples=1000, put=0.04, rate=100, sdrate=0.4, thresh=20)#How many of 1000 weeds of a genotype with resistance value 5.5 survive a herbicide application #with full dose? 'It is assumed that weeds reseaving less than 20 \% of the full dose survive #independently of their resistant value. intern_herbicide(resist=5.5, n_samples=1000, put=0.04, rate=100, sdrate=0.4, thresh=20)
struc_preparation2
Before the preparation (struc_preparation2) it is necessary to check, whether the necessary information is provided to principally conduct a simulation run. Furthermore the crop.list is brought to the correct form.
mod_check( param.weed = NA, area = NA, duration = NA, repetitions = NA, crop_list = NA, max_vec_length = NA )mod_check( param.weed = NA, area = NA, duration = NA, repetitions = NA, crop_list = NA, max_vec_length = NA )
param.weed |
A |
area |
number of area units. |
duration |
maximum number of simulation loops in the simulation. |
repetitions |
number of repetitions of the simulation. |
crop_list |
crop rotation. |
max_vec_length |
used internally, a technical term, defining the maximum length of vectors to be used. |
If no genetics are required af has to be set to NA. The value of dom is adjusted to the length of af. The dom is cut or the first given value is repeated to fit the number of given alleles in af. The variables af and dom are corrected and n_loci is created, which is 0 when no genetics is included.
mod_check(param.weed=param.GALAP, area=100, duration=2, repetitions=1, crop_list=c("corn"), max_vec_length=1000000)mod_check(param.weed=param.GALAP, area=100, duration=2, repetitions=1, crop_list=c("corn"), max_vec_length=1000000)
This function draws three figures. i)Numerics: mean number of individual plants of a specified development stage for each simulation cycle, summing up the genotypes to one number per cycle. ii)Raw counts: same as i) using the results of the repetitions instead of the mean. iii) Mean of the proportion of plants with only alleles for sensitivity, proportion of plants with only alleles for resistance, proportion of plants with mixed alleles for resistance and sensitivity, proportion of R alleles in the population.
## S4 method for signature 'prosper,missing' plot(x, y, plot_var = "SB_autumn", ...)## S4 method for signature 'prosper,missing' plot(x, y, plot_var = "SB_autumn", ...)
x |
|
y |
not used. |
plot_var |
variable, i.e. a column name of the result data.frame, to be plotted. |
... |
other graphical parameters. |
data(param.LOLRI) mod_lolri <- prosper.LOLRI(param.weed=param.LOLRI, area=20, af = c(0.005, 0.01), duration=3, repetitions=2) plot(mod_lolri)data(param.LOLRI) mod_lolri <- prosper.LOLRI(param.weed=param.LOLRI, area=20, af = c(0.005, 0.01), duration=3, repetitions=2) plot(mod_lolri)
pop_germc describes germination as random event for individual seeds. The function considers different cohorts and dormancy.
pop_germc(init_sb, germ, max_vec_length = 1e+07)pop_germc(init_sb, germ, max_vec_length = 1e+07)
init_sb |
column name of the initial seed bank in the data.frame dfgenotype that is delivered by struc_preparation. |
germ |
germination probabilities for the different cohorts. See details. |
max_vec_length |
used internally, a technical term, defining the maximum length of vectors to be used. |
Each individual has a chance to germinate or to stay dormant. In case of germination, it emerges in one of the cohorts. The distribution of individual seeds to cohorts or dormancy is random. The function uses the columns init_SB of the data.frame dfgenotype as input. The output values are the numbers of seedlings of each genotype and each cohort.
germ must be given as a numeric vector or - in case of multiple columns in init_sb - data.frame or matrix. In that case the row number must fit to the length of init_sb. If the columns in init_sb represent cohorts, the rows of germ give the germination probabilities for these specific cohorts. The sum of one row of germ shall be and . The difference of 1 and the sum of on row germ is the probability of dormancy.
Columns are added to dfgenotype: "germ_dorm" contains the numbers of each genotype that remain dormant, "germ1" to "germX" contain the numbers for each of X cohorts.
struc_preparation2(Rmx=10, af=c(0.01,0.8), epis=0, dom=1) ls() gen_freq( af=c(0.01,0.8), n_seeds=10000) #Distribute the individuals to three cohorts with the germination #probabilities 0.2, 0.4 and 0.4. pop_germc( init_sb="initialSB", germ=c(0.2,0.4,0.4)) rm(mf, dfgenotype, xprobab)struc_preparation2(Rmx=10, af=c(0.01,0.8), epis=0, dom=1) ls() gen_freq( af=c(0.01,0.8), n_seeds=10000) #Distribute the individuals to three cohorts with the germination #probabilities 0.2, 0.4 and 0.4. pop_germc( init_sb="initialSB", germ=c(0.2,0.4,0.4)) rm(mf, dfgenotype, xprobab)
Calculates the produced seeds and optionally the proportion of crop yield that is realized relative to weed free yield (Renton at al. 2011).
pop_reprod(start, area, kw, pen_co, kc, dc, crop_inr, SSmax, yield = FALSE)pop_reprod(start, area, kw, pen_co, kc, dc, crop_inr, SSmax, yield = FALSE)
start |
column names of parental cohorts. |
area |
number of area units. |
kw |
dimensionless weed competition parameter. |
pen_co |
penalty values for different weed cohorts. |
kc |
dimensionless crop competition parameter. |
dc |
crop sowing density, seeds per unit area. |
crop_inr |
position of the crop in the crop rotation. |
SSmax |
maximum of weed seed production per unit area. |
yield |
|
The number of produced seeds is calculated for 1m^2 by the formula:
The weed density dw is calculated for each squaremeter derived from the current simulation run (start). The used parameters values apply to wheat and ryegrass (Pannell et al. 2004, cited in Renton et al. 2011).
Renton, M.; Diggle, A.; Manalil, S. & Powles, S. (2011): Does cutting herbicide rates threaten the sustainability of weed management in cropping systems? Journal of Theoretical Biology, 283, 14-27. Pannell, D. J.; Stewart, V.; Bennett, A.; Monjardino, M.; Schmidt, C. & Powles, S. B. (2004): RIM: a bioeconomic model for integrated weed management of Lolium rigidum in Western Australia Agricultural Systems, 79, 305-325
struc_preparation2(Rmx=10, af=c(0.01,0.8), epis=0, dom=1) #Distribute 10000 individuals of the starting population across the genotypes provided by tmp. #The two gene loci have initial frequencies of 0.01 and 0.8. gen_freq(af=c(0.01,0.8), n_seeds=10000, max_vec_length=1e+07) pop_reprod("initialSB", area=100, kw=0.5, pen_co=1, kc=0.05, dc=100, crop_inr="wheat", SSmax=3000, yield=TRUE) rm(producedseeds, dfgenotype, mf, xprobab, propyield)struc_preparation2(Rmx=10, af=c(0.01,0.8), epis=0, dom=1) #Distribute 10000 individuals of the starting population across the genotypes provided by tmp. #The two gene loci have initial frequencies of 0.01 and 0.8. gen_freq(af=c(0.01,0.8), n_seeds=10000, max_vec_length=1e+07) pop_reprod("initialSB", area=100, kw=0.5, pen_co=1, kc=0.05, dc=100, crop_inr="wheat", SSmax=3000, yield=TRUE) rm(producedseeds, dfgenotype, mf, xprobab, propyield)
pop_step picks the individuals that will pass to the next development stage. This is a random process for every individual, which does not exert any selection pressure.
pop_step( start, start_comb = NA, result = NA, stepname = NA, surv_prob, max_vec_length = 1e+07 )pop_step( start, start_comb = NA, result = NA, stepname = NA, surv_prob, max_vec_length = 1e+07 )
start |
column names of parental cohorts. |
start_comb |
all named columns in |
result |
name of the results column. |
stepname |
name of the new column of dfgenotype added by this function. |
surv_prob |
probability to survive this step and reach the next growth stage. |
max_vec_length |
used internally, a technical term, defining the maximum length of vectors to be used. |
Individuals that reach the next growth stage are picked by using rbinom. In contrast to sel_herb, pop_step does not exert any evolutionary selection pressure.
When more than one column is selected with start, they are summed and the result is passed to the picking process. By setting start_comb the sum is added as a column to dfgenotype.
A new column is added to dfgenotype containing the surviving individuals of the different genotypes.
struc_preparation2(Rmx=10, af=c(0.01,0.8), epis=0, dom=1) gen_freq(af=c(0.01,0.8), n_seeds=10000) #How many individuals of each genotype will reach the next growth stage? pop_step(start="initialSB", stepname="survivingthewinter", surv_prob=0.4)struc_preparation2(Rmx=10, af=c(0.01,0.8), epis=0, dom=1) gen_freq(af=c(0.01,0.8), n_seeds=10000) #How many individuals of each genotype will reach the next growth stage? pop_step(start="initialSB", stepname="survivingthewinter", surv_prob=0.4)
PROSPER entails full parameterized models, which are described here.
Population dynamic models adress different purposes the models differ. The functions prosper.* present different adaptations.
prosper.ECHCG provides the setting for a simulation of the population dynamic of Echinochloa crus-galli.
prosper.GALAP provides the setting for a simulation of the population dynamic of Galium aparine. No selection process is used.
prosper.LOLRI performs a simulation of PROSPER using the setting presented by Renton at al. (2011). To manipulate the parameters see Details.
prosper.ECHCG( param.weed = PROSPER::param.ECHCG, area = NA, af = NA, dom = NA, epis = 0, put = 0.05, sdrate = 0.4, thresh = 20, Rmx = 10, rate = 100, duration = NA, repetitions = NA, crop_list = "corn", max_vec_length = 1e+07, undersowing = NA ) prosper.GALAP( param.weed = PROSPER::param.GALAP, sens_seeds = 400, area = 100, af = c(0.03, 0.08, 0.02), dom = c(0.5, 0.5, 0.5), epis = 0, put = 0.05, thresh = 20, Rmx = 10, rate = 100, sdrate = 0.4, duration = 15, repetitions = 1, crop_list = "wheat", max_vec_length = 1e+07 ) prosper.LOLRI( param.weed = PROSPER::param.LOLRI, area = 100, af = c(0.005, 0.01, 0.015, 0.02), dom = 0.5, epis = 0, put = 0.05, sdrate = 0.4, thresh = 20, Rmx = 10, dc = 150, kc = 1/11, kw = 1/33, SSmax = 30000, rate = 100, pen_co = c(1, 0.5), duration = 10, repetitions = 1, crop_list = c("wheat"), max_vec_length = 1e+07 )prosper.ECHCG( param.weed = PROSPER::param.ECHCG, area = NA, af = NA, dom = NA, epis = 0, put = 0.05, sdrate = 0.4, thresh = 20, Rmx = 10, rate = 100, duration = NA, repetitions = NA, crop_list = "corn", max_vec_length = 1e+07, undersowing = NA ) prosper.GALAP( param.weed = PROSPER::param.GALAP, sens_seeds = 400, area = 100, af = c(0.03, 0.08, 0.02), dom = c(0.5, 0.5, 0.5), epis = 0, put = 0.05, thresh = 20, Rmx = 10, rate = 100, sdrate = 0.4, duration = 15, repetitions = 1, crop_list = "wheat", max_vec_length = 1e+07 ) prosper.LOLRI( param.weed = PROSPER::param.LOLRI, area = 100, af = c(0.005, 0.01, 0.015, 0.02), dom = 0.5, epis = 0, put = 0.05, sdrate = 0.4, thresh = 20, Rmx = 10, dc = 150, kc = 1/11, kw = 1/33, SSmax = 30000, rate = 100, pen_co = c(1, 0.5), duration = 10, repetitions = 1, crop_list = c("wheat"), max_vec_length = 1e+07 )
param.weed |
A |
area |
number of area units. |
af |
initial frequency of resistance alleles in the population. |
dom |
dominance of resistance alleles. |
epis |
epistasis value, describing the interaction between resistance alleles. |
put |
probability of a weed to be untouched by the herbicide. |
sdrate |
variance of the herbicide rate reaching the weed. |
thresh |
threshold herbicide rate to kill weeds without resistance. |
Rmx |
maximum resistance value, if all gene loci under consideration are homozygous resistant. |
rate |
percentage (%) of the registered herbicide dose. |
duration |
maximum number of simulation loops in the simulation. |
repetitions |
number of repetitions of the simulation. |
crop_list |
crop rotation. |
max_vec_length |
used internally, a technical term, defining the maximum length of vectors to be used. |
undersowing |
Numerical vector with two values between 0 and 1. See details. |
sens_seeds |
sensitive seeds added every year. |
dc |
crop sowing density, seeds per unit area. |
kc |
dimensionless crop competition parameter. |
kw |
dimensionless weed competition parameter. |
SSmax |
maximum of weed seed production per unit area. |
pen_co |
penalty values for different weed cohorts. |
prosper.ECHCG() simulates originally the population dynamic of Echinochloa crus-galli using the data param.ECHCG. Different cohorts of weed seedlings are the focus of this model. The focus of this model is the effect of weeds that escape the selection pressure of herbicide treatment. These weeds keep the unselected genetic. Can they buffer the selection process? E. crus-galli is able to germinate over a long period after maize planting with decreasing reproductive success (Bagavathiannan, 2013). In the model all germinating individuals are represented by two cohorts; an early, major cohort with with high seed production, and a small late emerging with lower reproduction. Only the first cohort is controlled by a herbicide, which is a typical situation in Germany (Rossberg, 2016). The second cohort escapes the herbicide treatment unaffected. However, the second cohort can be suppressed, for example by an undersown crop. Three scenarios with different degrees of suppression, 0%, 30% and 100%, were simulated (Redwitz, 2016).
The parameter undersowing describes the probability of surviving a second, not selective pressure on weed seedlings, which germinate after the selective herbicide was applied.
prosper.ECHCG provides the setting for a simulation of the population dynamic of Echinochloa crus-galli.
prosper.GALAP() simulates originally the population dynamic of Galium aparine using the data param.GALAP. Whether sowing of susceptible weed seeds can restore an 'acceptable' resistance level of a population in the early stages of resistance development, is an extraordinary research question. The patchy occurrence of Galium aparine and its large seeds result in highly variable population dynamic parameters. Modeling has to take into account this variability. We used a simple population dynamics model structure (Redwitz et al., 2015). A seedbank in spring provides seeds out of which one cohort is germinating. The weeds are selected by herbicides, produce seeds, which are affected by seed predation and return to the seedbank in autumn. Data of a long term field experiment were used for parametrization (Daedlow, 2015).
prosper.LOLRI() performs a simulation of Lolium rigidum similar to PERTH (Renton et al. 2011) when it is used with param.LOLRI.
prosper.ECHCG: Population dynamic model of Echinochloa crus-galli
prosper.GALAP: Population dynamic model of Galium aparine
prosper.LOLRI: Population dynamic model of Lolium multiflorum
Christoph von Redwitz, [email protected]
Redwitz, C von, Pannwitt H, Gerowitt B (2016): About the interplay of sensitive and resistant biotypes in weed populations - simulation exercises for Echinochloa crus-galli in maize crops. Proceedings - 28th German Conference on Weed Biology and Weed Control, Julius-Kuehn-Archiv, 93-99, 452.
Redwitz, C von, Daedlow D, Gerowitt B (2015): Simulation exercises on long-term management of widespread herbicide resistance in a field weed population. Proceedings 17th Symposium of the European Weed Research Society, Montpellier, France, 108.
Renton, M., Diggle, A., Manalil, S. & Powles, S. (2011): Does cutting herbicide rates threaten the sustainability of weed management in cropping systems? Journal of Theoretical Biology, Elsevier BV, 283, 14-27.
## Not run: mod_echcg <- prosper.ECHCG(param.weed = param.ECHCG, area=100, af=c(0.001), undersowing=0.2,dom=0.5,duration=7,repetitions=1) #The model call for Redwitz et al. (2015) undersowing_prob <- c(1, 0.3, 0) #no undersowing, strong competition, complete dominance years <- 20 reps <- 4 ####------------------------ simu_collect <- list() for(simu in 1:3){ simu_collect[[simu]] <- prosper.ECHCG(area = 100, param.weed = param.ECHCG, thresh = 20, duration = years, af = 0.001, dom = 1, undersowing = undersowing_prob[simu], repetitions = reps ) } ## End(Not run) mod_galap <- prosper.GALAP(param.weed=param.GALAP, repetitions=2, duration=10) mod_lolri <- prosper.LOLRI(param.weed=param.LOLRI, area=100, duration=15, repetitions=3)## Not run: mod_echcg <- prosper.ECHCG(param.weed = param.ECHCG, area=100, af=c(0.001), undersowing=0.2,dom=0.5,duration=7,repetitions=1) #The model call for Redwitz et al. (2015) undersowing_prob <- c(1, 0.3, 0) #no undersowing, strong competition, complete dominance years <- 20 reps <- 4 ####------------------------ simu_collect <- list() for(simu in 1:3){ simu_collect[[simu]] <- prosper.ECHCG(area = 100, param.weed = param.ECHCG, thresh = 20, duration = years, af = 0.001, dom = 1, undersowing = undersowing_prob[simu], repetitions = reps ) } ## End(Not run) mod_galap <- prosper.GALAP(param.weed=param.GALAP, repetitions=2, duration=10) mod_lolri <- prosper.LOLRI(param.weed=param.LOLRI, area=100, duration=15, repetitions=3)
PROSPER simulates the development of annual plant populations with regard to population dynamics and genetics, especially herbicide resistance. It combines genetics on the individual level with a stochastic development on the population level. The genetic part is modeled after Renton et al. (2011), the stochastic part follows Daedlow (2015). All parameters delivered with PROSPER are based on an area of 1 squaremeter.
Renton, M.; Diggle, A.; Manalil, S. & Powles, S. (2011): Does cutting herbicide rates threaten the sustainability of weed management in cropping systems? Journal of Theoretical Biology, 283, 14-27.
Daedlow, D. (2015): About the contribution of seed predation on weed demography. (Doctoral dissertation). University of Rostock, Faculty of Agriculture and Environmental Sciences. GBV Gemeinsamer Verbundkatalog (Accession No. 839752644).
This is a working horse of PROSPER. It quantifies the number or the proportion of individuals entering the next development stage using predefined formulas (formul). Typically these formulas are the results of experiments.
quanti( origin = NA, step_name, crop, proportion = TRUE, equal_dis = TRUE, res_max = NA, res_min = 0, formul = NA, area, addit_variables = NA, log_values = TRUE, back_log = TRUE )quanti( origin = NA, step_name, crop, proportion = TRUE, equal_dis = TRUE, res_max = NA, res_min = 0, formul = NA, area, addit_variables = NA, log_values = TRUE, back_log = TRUE )
origin |
numbers or log-numbers of individuals at the start for every genotype. |
step_name |
step name in the database containing the model parameters. |
crop |
current crop in the crop sequence, must fit to one of the crop names in the data frame with weed-parameters. |
proportion |
|
equal_dis |
|
res_max |
output maximum, ignored when |
res_min |
output minimum. |
formul |
|
area |
number of area units. |
addit_variables |
variables used in |
log_values |
logical, |
back_log |
logical, |
Within PROSPER simulation models are build up with discrete simulation steps. These steps are conducted by functions like pop_step or gen_reprod. These functions affect the complete population and need a count of individuals or a proportion of the population that are affected of the specific simulation step. These numbers are calculated with quanti. The calculation is based on the data provided in weed-parameters. There a model is given for every parameter in the table. The parameter formul allows to use a different model if necessary. If no model is given at all, the simulation step is assumed to consist only of one value. The parameters of param.weed are normal distributed and the SE is used to draw them for the current calculation using rnorm(). These values are used to evaluate the model for simulation step. The resulting count or proportion can be used to perform the simulation step for the population.
Either a (log-)number of individuals or a proportion resp. a rate.
pop_step pop_germc weed-parameters
#loads the example data for Echinochloa crus-galli data(param.ECHCG) param.weed <- param.ECHCG #how many seeds (natural, not log-scale) prodused by 100 plants in a corn stand on 100 area units? quanti(origin=100, step_name="seed_prod_first", crop="corn", proportion=FALSE, area=100, log_values=FALSE, back_log=FALSE) rm(param.ECHCG)#loads the example data for Echinochloa crus-galli data(param.ECHCG) param.weed <- param.ECHCG #how many seeds (natural, not log-scale) prodused by 100 plants in a corn stand on 100 area units? quanti(origin=100, step_name="seed_prod_first", crop="corn", proportion=FALSE, area=100, log_values=FALSE, back_log=FALSE) rm(param.ECHCG)
sel_herb calculates the surviving number of each genotype. sel_herb selects for resistant individuals.
sel_herb(start, result, thresh, sdrate, rate, put, max_vec_length = 1e+07)sel_herb(start, result, thresh, sdrate, rate, put, max_vec_length = 1e+07)
start |
column names of parental cohorts. |
result |
name of the results column. |
thresh |
threshold herbicide rate to kill weeds without resistance. |
sdrate |
variance of the herbicide rate reaching the weed. |
rate |
percentage (%) of the registered herbicide dose. |
put |
probability of a weed to be untouched by the herbicide. |
max_vec_length |
used internally, a technical term, defining the maximum length of vectors to be used. |
For every genotype intern_herbicide is called. If no genetics are included, the value from start is returned in result.
struc_preparation2(Rmx=10, af=c(0.02,0.01), epis=0, dom=1) gen_freq( af=c(0.01,0.8), n_seeds=10000) sel_herb(start="initialSB", result="winter", thresh=20, sdrate=0.4, rate=100, put=0.04)struc_preparation2(Rmx=10, af=c(0.02,0.01), epis=0, dom=1) gen_freq( af=c(0.01,0.8), n_seeds=10000) sel_herb(start="initialSB", result="winter", thresh=20, sdrate=0.4, rate=100, put=0.04)
Calculates the phenotypic resistance value (Renton et al. 2011) for each genotype in dfgenotype.
sel_resist(Rmx, epis, dom)sel_resist(Rmx, epis, dom)
Rmx |
maximum resistance value, if all gene loci under consideration are homozygous resistant. |
epis |
epistasis value, describing the interaction between resistance alleles. |
dom |
dominance of resistance alleles. |
This function is used in struc_preparation2 to calculate the resistance value using the following term:
Resistance values range from 1 to Rmx. The higher the resistance value is, the less the plant is susceptible to the herbicide.
sel_resist is only used by struc_preparation2.
numeric vector with resistance values
Renton, M.; Diggle, A.; Manalil, S. & Powles, S. (2011): Does cutting herbicide rates threaten the sustainability of weed management in cropping systems? Journal of Theoretical Biology, 283, 14-27.
#dfgenotype is usually generated by the function 'struc_preparation2'. #Here, a simple example is done by hand. var1 <- c("00","01","02","10","11","12","20","21","21") var2 <- c(0,0,0,1,1,1,2,2,2) var3 <- c(0,1,2,0,1,2,0,1,2) dfgenotype <- data.frame(genotype=var1,l1=var2,l2=var3, stringsAsFactors = TRUE) sel_resist(Rmx=10, epis=0, dom=1)#dfgenotype is usually generated by the function 'struc_preparation2'. #Here, a simple example is done by hand. var1 <- c("00","01","02","10","11","12","20","21","21") var2 <- c(0,0,0,1,1,1,2,2,2) var3 <- c(0,1,2,0,1,2,0,1,2) dfgenotype <- data.frame(genotype=var1,l1=var2,l2=var3, stringsAsFactors = TRUE) sel_resist(Rmx=10, epis=0, dom=1)
Conditions that are necessary to continue the simulation are checked.
struc_endSim(simcycle = year, break_col_names = "SB_autumn")struc_endSim(simcycle = year, break_col_names = "SB_autumn")
simcycle |
The count of simulation runs. The value is |
break_col_names |
The names of the columns in |
This function combines two conditions to terminate the simulation. The first is the duration that is defined for the simulation. The second is the the number of individuals for the next simulation cycle: when the population is extinct the simulation ends.
logical TRUE if the simulation has to stop.
struc_preparation creates the data input and output structures (data.frames and table) for the simulation run, 'dfgenotype', 'xprobab' and 'mf'.
struc_preparation2(Rmx = NA, af = NA, epis = NA, dom = NA)struc_preparation2(Rmx = NA, af = NA, epis = NA, dom = NA)
Rmx |
maximum resistance value, if all gene loci under consideration are homozygous resistant. |
af |
initial frequency of resistance alleles in the population. |
epis |
epistasis value, describing the interaction between resistance alleles. |
dom |
dominance of resistance alleles. |
Prior to the simulation, a data.frame is generated to save results (dfgenotype). Additionally, a table with recombination probabilities (xprobab) is calculated. During the simulation run, probability values are not computed again but looked up in the table. PROSPER assumes diploid plants and maximum four resistance genes. To calculate the phenotypic resistance value for each genotype sel_resist is called. If n_loci < 0 the structure is set up for no genetics at all in the simulation.
Returns a list of two data.frame and a table:
mf: all possible combinations of parental genotypes (see 'dfgenotype$genotype') are saved in one column 'mf' (male, female). The column 'mf' is a character vector. Each string of the vector has twice the length of the number of resistance loci under consideration.
dfgenotype: the structure to save the results of one simulation cycle (year). After each cycle the data is reset to the new start values. The first column 'genotype' is a character vector. Each string of the vector has the length of the number of resistance loci under consideration. Each locus can have 0, 1 or 2 resistance alleles. The second column 'resist' saves resistance values that are calculated according to the equation in the section 'details'.
xprobab: the probabilities of occurrence for all possible genotypes in the offspring (F-generation) with all possible parent genotypes (P-generation). Free recombination is assumed. Column names are the combinations of parental genes, row names are genotypes of the offspring.
The run of struc_preparation is time consuming. Duration strongly increases with the number of genes under consideration, n_loci.
#generate the genotype and probability tables for a simulation with two resistance #loci with one dominant and one partial dominant resistant allele, no epistasis, and a #maximumx resistance value of 10. ls() struc_preparation2(Rmx=10, af=c(0.01,0.5), epis=0, dom=c(1,0.3)) ls() rm(dfgenotype, mf, xprobab)#generate the genotype and probability tables for a simulation with two resistance #loci with one dominant and one partial dominant resistant allele, no epistasis, and a #maximumx resistance value of 10. ls() struc_preparation2(Rmx=10, af=c(0.01,0.5), epis=0, dom=c(1,0.3)) ls() rm(dfgenotype, mf, xprobab)
struc_saveSimData collects the data of one simulation cycle and returns an object to start a new simulation cycle.
struc_saveSimData(rep_counter, simcycle, start_names, end_names, simstruc)struc_saveSimData(rep_counter, simcycle, start_names, end_names, simstruc)
rep_counter |
number of the current repetition. |
simcycle |
number of the current simulation cycle. |
start_names |
names of the columns with the first population stage in the simulation. |
end_names |
names of the columns with the last population stage in the simulation. These are taken as the first stage in the next simulation cycle. |
simstruc |
two numbers, the first defines the number of repetitions for the simulation, the second defines the number of simulation cycles in each repetition of the simulation. |
In the first simulation cycle the object sim_result of the class prosper is created and the first cycle results are saved. This object is the same for all cycles in all repetitions and gets all the results. The function can only be called once per simulation run.
A prosper object. SimData as a data.table with the repetition in the first and the simulation cycle in the second column.
struc_preparation2(Rmx=10, af=c(2,1), epis=0, dom=c(1,0.3)) dfgenotype$"SB_autumn" <- c(4,23,0,123,53,98,45,3245,234) dfgenotype$"SB_autumn_end" <- c(0,2,0,123,434,5234,5678,123,2) #creation of an example object with data of the first year struc_saveSimData(rep_counter=1, simcycle=1, simstruc=c(5, 10), start_names="SB_autumn", end_names="SB_autumn_end") #creating some fictional population stages dfgenotype$"SB_autumn" <- c(1,1,1,1,0,0,0,0,4) dfgenotype$"SB_autumn_end" <- c(67,67,67,67,67,67,67,67,67) #appending rows with the new results to the first results. necessary columns are inserted. struc_saveSimData( rep_counter=1, simcycle=2, start_names="SB_autumn", end_names="SB_autumn_end", simstruc=c(repetitions, duration)) sim_result rm(sim_result, dfgenotype, mf, xprobab)struc_preparation2(Rmx=10, af=c(2,1), epis=0, dom=c(1,0.3)) dfgenotype$"SB_autumn" <- c(4,23,0,123,53,98,45,3245,234) dfgenotype$"SB_autumn_end" <- c(0,2,0,123,434,5234,5678,123,2) #creation of an example object with data of the first year struc_saveSimData(rep_counter=1, simcycle=1, simstruc=c(5, 10), start_names="SB_autumn", end_names="SB_autumn_end") #creating some fictional population stages dfgenotype$"SB_autumn" <- c(1,1,1,1,0,0,0,0,4) dfgenotype$"SB_autumn_end" <- c(67,67,67,67,67,67,67,67,67) #appending rows with the new results to the first results. necessary columns are inserted. struc_saveSimData( rep_counter=1, simcycle=2, start_names="SB_autumn", end_names="SB_autumn_end", simstruc=c(repetitions, duration)) sim_result rm(sim_result, dfgenotype, mf, xprobab)
This function gives a first overview of a prosper simulation result.
## S4 method for signature 'prosper' summary(object, geneticSumCol = "SB_autumn_end", ...)## S4 method for signature 'prosper' summary(object, geneticSumCol = "SB_autumn_end", ...)
object |
|
geneticSumCol |
column name of the prosper object the genetic summary is built of. |
... |
not in use. |
For the population dynamic part, means and standard deviations per simulation cycle for every numeric variable is calculated. Variability is introduced by repetitions. Additionally, a simple overview on factorial variables is given. For further calculations the numbers of individuals for each simulation cycle and repetition are provided.
For the genetic part, means and standard deviations of proportions of individuals with only sensitive alleles (allSalleles), only resistance alleles (allRalleles), mixed alleles (RSalleles), and the proportions of resistance alleles in the population (Ralleles) are calculated for each simulation cycle. The repetitions are responsible for the variability.
A named list of data.tables.
data(param.LOLRI) mod_lolri <- prosper.LOLRI(param.weed=param.LOLRI, area=30, af = c(0.005, 0.01), duration=3, repetitions=2) summary(mod_lolri)data(param.LOLRI) mod_lolri <- prosper.LOLRI(param.weed=param.LOLRI, area=30, af = c(0.005, 0.01), duration=3, repetitions=2) summary(mod_lolri)
PROSPER entails full parameterized models. The parameterisation for these models are described here.
The provided data are given in tables with 10 columns.
the weed species.
the crop for which the parameters apply, character.
the name of the variable, simulation step, character.
a naming of parameter, like a variable definition, character.
the mean of the parameters defined with "name", numeric.
the standard error of the parameters defined with "name", numeric.
a (statistical) model to describe a specific simulation step. The model must be constructed of defined parameters with "name", character.
a model might have a minimum level of use, then this value can be defined here, numeric.
a model might have a maximum level of use, then this value can be defined here, numeric.
if the value is taken from sources not owned by the author they can be named here.
param.LOLRI entails the data used for the introduction of PERTH by Renton et al. (2011) using the parameters of Pannell et al. (2004). With prosper.LOLRI one of the PERTH models can be reproduced.
param.ECHCG contains parameters for population dynamic of Echinochloa crus-galli (Redwitz et al., 2016).
param.GALAP contains parameters for population dynamic off Galium aparine taken from field experiments (Daedlow, 2015).
Bagavathiannan, MV, Norsworthy, JK (2013) Postdispersal loss of important arable weed seeds in the midsouthern United States. Weed Science, 61(4), 570-579.
Bosnic, AC, Swanton, C (1997) Influence of Barnyardgrass (Echinochloa crus-galli) Time of Emergence and Density on Corn (Zea mays). Weed Science, 45(2), 276-282.
Clay, SA, Kleinjan, J, Clay, DE, Forcella, Frank, Batchelor, W (2005) Growth and fecundity of several weed species in corn and soybean. Agronomy journal, 97(1), 294-302.
Daedlow, D. (2015): About the contribution of seed predation on weed demography. (Doctoral dissertation). University of Rostock, Faculty of Agriculture and Environmental Sciences. GBV Gemeinsamer Verbundkatalog (Accession No. 839752644).
Ogg, AG, Dawson, JH (1984) Time of emergence of eight weed species. Weed Science, 32(3), 327-335.
Pannell, DJ, Stewart, V, Bennett, A, Monjardino, M, Schmidt, C, Powles, SB (2004) RIM:a bioeconomic model for integrated weedmanagement of Lolium rigidum in Western Australia. Agricultural Systems, 79, 305-325.
Redwitz, C von, Pannwitt, H, Gerowitt, B (2016): About the interplay of sensitive and resistant biotypes in weed populations - simulation exercises for Echinochloa crus-galli in maize crops. Julius-Kuehn-Archiv 452, 93-99.
Renton, M.; Diggle, A.; Manalil, S. & Powles, S. (2011): Does cutting herbicide rates threaten the sustainability of weed management in cropping systems? Journal of Theoretical Biology, 283, 14-27.