Bstep: Bstep Function to increment haddock biomass 1 year

In rBatt/fishyWarnings: How should management respond to warning signals of haddock collapse?

Description

Bstep Function to increment haddock biomass 1 year

Usage

Bstep(B, qE = 0.15, sdU = 0.1, dt = 0.1, fB = c("SH", "schaefer"), ...)

Arguments

B	biomass in kt
qE	catchability*effort = F (harvest rate); F*B = C, catch
sdU	standard deviation of the random shock; set to 0 for deterministic simulation
dt	within-season size of time step; should be a number > 0 and <= 1
fB	function of B; should be a character of either "SH" or "schaefer", corresponding to one of the two models
...	arguments passed to fB

Details

Follows equation 1 in Spencer and Collie 1997.

Note

In the example below, the equilibrium value (average between 2020 and 2040) for F=0.18 is different under low-noise and high-noise scenarios. Below the simulation is as in the paper, but if the noise level is decreased (or if simulated deterministically), then the equiblirum is much closer to the cases of F=0.21 and F=0.24. I think this is a noise-induced transition. Cool!

Examples

# Make Figure 5a
nY <- 50
# 13.8 kt is from pg 2922, second column, penultimate paragraph; 
# 13.8 is the starting value in 1993
makeB <- function(){c(13.8, rep(NA, nY-1))}
Year <- seq(1993, length.out=nY)
qE <- c(0, 0.06, 0.12, 0.18, 0.21, 0.24)
Bvec <- replicate(length(qE), makeB())
for(j in 1:length(qE)){
	for(i in 2:nrow(Bvec)){
		Bvec[i,j] <- Bstep(Bvec[i-1,j], qE[j], sdU=0.1)
	}
}
ltys <- c("solid", "dotted", "dashed", "longdash", "twodash", "dotdash")
ylim <- range(Bvec, na.rm=TRUE)
plot(Year, Bvec[,1], type='l', 
	lty=ltys[1], ylim=ylim, 
	ylab="Biomass (kt)", xlab="time", main="STH model"
)
for(j in 2:ncol(Bvec)){
	lines(Year, Bvec[,j], lty=ltys[j])
}
legend("topleft", paste("F", qE, sep="="), lty=ltys)

rBatt/fishyWarnings documentation built on May 26, 2019, 7:45 p.m.