fishmethods (version 1.12-1)

plot.grotagplus: Plotting Tagging-Growth Objects

Description

Plotting method for output from function grotagplus, which has class "grotagplus".

Usage

# S3 method for grotagplus
plot(x,plot.type="meangrowth",Linitial=NULL,resid.spec=list(Pearson=T,
      x="mean.delL"),xlim=NULL,ylim=NULL,pch=20,leg.loc=NULL,
      age.based.growth=NULL,...)

Arguments

x

Growth-model fit to tagging data as output by function "grotagplus".

plot.type

Character string identifying the type of plot required: "meangrowth" = mean annual growth vs initial length; "traj" = one-year growth trajectory of fish of initial length specified by Linitial; or "resid" = plot of ordinary or Pearson residuals (plot details specified by resid.spec).

Linitial

Initial length to use for plot of growth trajectory.

resid.spec

List, specifying details of a residual plot, with components "Pearson" (logical, if T [default] plot Pearson residuals, otherwise simple residuals) and "x" (the x-variable in the plot - either "L1", length at tagging; "delT", time at liberty; or "mean.delL", expected length increment).

xlim

Allow the user to set x-limits for a plot that differ from those defined by the range of the plotted data.

ylim

Allow the user to set y-limits for a plot that differ from those defined by the range of the plotted data.

pch

Allows the user to change the plotting symbol for residual plots from the default pch=20.

leg.loc

Allows the user to change the legend location from its default position ("topright" for meangrowth and resid; "topleft" for traj). Note that a legend is used only for traj or for other plots with multiple datasets.

age.based.growth

This argument allows the user to add, to a meangrowth plot, growth estimates (plotted as dashed lines) from age-length datasets. It should be a list of vectors, each of which contains estimates of mean length corresponding to a vector of increasing ages whose increments are always 1 year (the ages are not included in the argument because they are not used in the plot, and the age vectors need not be the same in each component). If the list is named then the names will be interpreted as identifying different datasets. If a name appears in fit$datasetnames the age-based growth will be plotted with the same colour as the corresponding tagging growth. If the list is not named then it must be of the same length as fit$datasetnames (or of length 1 if there is only one dataset in the tagging data) and it will be assumed that the ith component corresponds to the ith tagging dataset.

...

Other graphical parameters. See par

Author

Chris Francis

Gary A. Nelson, Massachusetts Division of Marine Fisheries

Marco Kienzle

Details

Examples of the three plot types are given in Figs 7 & 8 of Francis and Francis (1992), for "resid" and "meangrowth", respectively; and in Fig. 2 of Francis (1988), for "traj".

plot.type="meangrowth" is the recommended way for plotting growth rates estimated from tagging data. Argument age.based.growth allows a rough comparison between these growth estimates and those from age-length data (the comparison is between the mean growth at length L and that at the age for which the mean length is L).

The traj plot, as well as showing the mean (i.e., expected) growth (solid line), shows 95 (dashed lines) and with (dotted lines) allowance for measurement error.

In residual plots, a dashed lowess line is plotted for each dataset to indicate any trend and, for Pearson residuals, dotted lines at +/- 2 indicate approximate 95

For fits using multiple datasets, colour is used to distinguish the datasets. Use "palette" to change the match between colour and dataset (the ith colour in the palette is associated with the ith element in fit$datasetnames).

References

1 Francis, R.I.C.C., 1988. Maximum likelihood estimation of growth and growth variability from tagging data. New Zealand Journal of Marine and Freshwater Research, 22, p.42-51.

2 Francis, M.P. and Francis, R.I.C.C. 1992. Growth rate estimates for New Zealand rig (Mustelus lenticulatus). Australian Journal of Marine and Freshwater Research 43: 1157-1176

See Also

grotagplus print.grotagplus

Examples

# Plot of mean growth like that in Fig 8. of Francis & Francis (1992)
data(rig)
fit <- grotagplus(rig,dataID="Sex",alpha=70,beta=100,
                  model=list(mean="Francis",var="linear",seas="none"),
                 design=list(galpha=list("F","M"),gbeta=list("F","M"),
                             s=1,nu=1,m=0,p=0),
                  stvalue=list(galpha=c(5,4),gbeta=c(3,2),s=2,nu=0.5),
                  upper=list(galpha=c(8,6),gbeta=c(5,4),s=4,nu=1),
                 lower=list(galpha=c(3,2),gbeta=c(1.5,1),s=0.5,nu=0.2))
mnlenatage <- list(F=90.7*(1-exp(-0.42*(seq(1.5,6.5)-0.77))),
           M= 118.7*(1-exp(-0.16*(seq(4,11)-2.02))),
           PGM=161.1*(1-exp(-0.11*(seq(3.5,10.5)-1.91))))
plot(fit,age.based.growth=mnlenatage)
## Residual plots
fit <- grotagplus(rig,dataID="Sex",alpha=70,beta=100,
                  model=list(mean="Francis",var="linear",seas="none"),
                 design=list(galpha=list("F","M"),gbeta=list("F","M"),
                             s=1,nu=1,m=0,p=0),
                  stvalue=list(galpha=c(5,4),gbeta=c(3,2),s=2,nu=0.5),
                  upper=list(galpha=c(8,6),gbeta=c(5,4),s=4,nu=1),
                 lower=list(galpha=c(3,2),gbeta=c(1.5,1),s=0.5,nu=0.2))
plot(fit,"resid")
plot(fit,"resid",resid.spec=list(Pearson=FALSE,x="L1"))
## Trajectory plot as in Fig. 2 of Francis (1988)
data(bonito)
fit <- grotagplus(bonito,alpha=35,beta=55,
               design=list(galpha=1,gbeta=1,s=1,nu=1,m=1,p=1,u=1,w=1),
               stvalue=list(s=0.81,nu=0.3,m=0,p=0.01,u=0.5,w=0.5),
               upper=list(s=3,nu=1,m=2,p=0.1,u=1,w=1),
               lower=list(s=0.1,nu=0.1,m=-2,p=0,u=0,w=0))
plot(fit,"traj",Linitial=35)