Learn R Programming

FSA (version 0.8.11)

growthModels: Creates a function for a specific parameterization of the von Bertalanffy, Gompertz, Richards, and logistic growth functions.

Description

Creates a function for a specific parameterizations of the von Bertalanffy, Gompertz, Richards, and logistic growth functions. Use growthFunShow() to see the equations for each growth function.

Usage

vbFuns(param = c("Typical", "typical", "Traditional", "traditional", "BevertonHolt", "Original", "original", "vonBertalanffy", "GQ", "GallucciQuinn", "Mooij", "Weisberg", "Ogle", "Schnute", "Francis", "Laslett", "Polacheck", "Somers", "Somers2", "Pauly", "Fabens", "Fabens2", "Wang", "Wang2", "Wang3"), simple = FALSE, msg = FALSE)
GompertzFuns(param = c("Ricker1", "Ricker2", "Ricker3", "QuinnDeriso1", "QuinnDeriso2", "QuinnDeriso3", "QD1", "QD2", "QD3", "Original", "original", "Troynikov1", "Troynikov2"), simple = FALSE, msg = FALSE)
RichardsFuns(param = 1, simple = FALSE, msg = FALSE)
logisticFuns(param = c("CJ1", "CJ2", "Karkach", "Haddon", "CampanaJones1", "CampanaJones2"), simple = FALSE, msg = FALSE)
growthFunShow(type = c("vonBertalanffy", "Gompertz", "Richards", "Logistic", "Schnute"), param = NULL, case = param, plot = FALSE, ...)

Arguments

param
A string (for von Bertalanffy, Gompertz, and logistic) or numeric (for Richards) that indicates the specific parameterization of the growth function. See details.
simple
A logical that indicates whether the function will accept all parameter values in the first parameter argument (=FALSE; DEFAULT) or whether all individual parameters must be specified in separate arguments (=TRUE).
msg
A logical that indicates whether a message about the growth function and parameter definitions should be output (=TRUE) or not (=FALSE; DEFAULT).
type
A string (in growthFunShow) that indicates the type of growth function to show.
case
A numeric that indicates the specific case of the Schnute function to use. See details.
plot
A logical that indicates whether the growth function expression should be shown as an equation in a simple plot.
...
Not implemented.

Value

The functions ending in xxxFuns return a function that can be used to predict fish size given a vector of ages and values for the growth function parameters and, in some parameterizations, values for constants. The result should be saved to an object that is then the function name. When the resulting function is used, the parameters are ordered as shown when the definitions of the parameters are printed after the function is called (if msg=TRUE). If simple=FALSE (DEFAULT), then the values for all parameters may be included as a vector in the first parameter argument (but in the same order). Similarly, the values for all constants may be included as a vector in the first constant argument (i.e., t1). If simple=TRUE, then all parameters and constants must be declared individually. The resulting function is somewhat easier to read when simple=TRUE, but is less general for some applications.An expression of the equation for each growth function may be created with growthFunShow. In this function type= is used to select the major function type (e.g., von Bertalanffy, Gompertz, Richards, Logistic, Schnute) and param= is used to select a specific parameterization of that growth function. If plot=TRUE, then a simple graphic will be created with the equation using plotmath for a pretty format.

IFAR Chapter

12-Individual Growth.

References

Ogle, D.H. 2016. Introductory Fisheries Analyses with R. Chapman & Hall/CRC, Boca Raton, FL.

Campana, S.E. and C.M. Jones. 1992. Analysis of otolith microstructure data. Pages 73-100 In D.K. Stevenson and S.E. Campana, editors. Otolith microstructure examination and analysis. Canadian Special Publication of Fisheries and Aquatic Sciences 117. [Was (is?) from http://www.dfo-mpo.gc.ca/Library/141734.pdf.]

Fabens, A. 1965. Properties and fitting of the von Bertalanffy growth curve. Growth 29:265-289.

Francis, R.I.C.C. 1988. Are growth parameters estimated from tagging and age-length data comparable? Canadian Journal of Fisheries and Aquatic Sciences, 45:936-942.

Gallucci, V.F. and T.J. Quinn II. 1979. Reparameterizing, fitting, and testing a simple growth model. Transactions of the American Fisheries Society, 108:14-25.

Garcia-Berthou, E., G. Carmona-Catot, R. Merciai, and D.H. Ogle. A technical note on seasonal growth models. Reviews in Fish Biology and Fisheries 22:635-640. [Was (is?) from https://www.researchgate.net/publication/257658359_A_technical_note_on_seasonal_growth_models.]

Gompertz, B. 1825. On the nature of the function expressive of the law of human mortality, and on a new method of determining the value of life contingencies. Philosophical Transactions of the Royal Society of London. 115:513-583.

Haddon, M., C. Mundy, and D. Tarbath. 2008. Using an inverse-logistic model to describe growth increments of blacklip abalone (Haliotis rubra) in Tasmania. Fishery Bulletin 106:58-71. [Was (is?) from http://aquaticcommons.org/8857/1/haddon_Fish_Bull_2008.pdf.]

Karkach, A. S. 2006. Trajectories and models of individual growth. Demographic Research 15:347-400. [Was (is?) from http://www.demographic-research.org/volumes/vol15/12/15-12.pdf.]

Katsanevakis, S. and C.D. Maravelias. 2008. Modelling fish growth: multi-model inference as a better alternative to a priori using von Bertalanffy equation. Fish and Fisheries 9:178-187.

Mooij, W.M., J.M. Van Rooij, and S. Wijnhoven. 1999. Analysis and comparison of fish growth from small samples of length-at-age data: Detection of sexual dimorphism in Eurasian perch as an example. Transactions of the American Fisheries Society 128:483-490.

Polacheck, T., J.P. Eveson, and G.M. Laslett. 2004. Increase in growth rates of southern bluefin tuna (Thunnus maccoyii) over four decades: 1960 to 2000. Canadian Journal of Fisheries and Aquatic Sciences, 61:307-322.

Quinn, T. J. and R. B. Deriso. 1999. Quantitative Fish Dynamics. Oxford University Press, New York, New York. 542 pages.

Quist, M.C., M.A. Pegg, and D.R. DeVries. 2012. Age and Growth. Chapter 15 in A.V. Zale, D.L Parrish, and T.M. Sutton, Editors Fisheries Techniques, Third Edition. American Fisheries Society, Bethesda, MD.

Richards, F. J. 1959. A flexible growth function for empirical use. Journal of Experimental Biology 10:290-300.

Ricker, W.E. 1975. Computation and interpretation of biological statistics of fish populations. Technical Report Bulletin 191, Bulletin of the Fisheries Research Board of Canada. [Was (is?) from http://www.dfo-mpo.gc.ca/Library/1485.pdf.]

Ricker, W.E. 1979. Growth rates and models. Pages 677-743 In W.S. Hoar, D.J. Randall, and J.R. Brett, editors. Fish Physiology, Vol. 8: Bioenergetics and Growth. Academic Press, NY, NY. [Was (is?) from https://books.google.com/books?id=CB1qu2VbKwQC&pg=PA705&lpg=PA705&dq=Gompertz+fish&source=bl&ots=y34lhFP4IU&sig=EM_DGEQMPGIn_DlgTcGIi_wbItE&hl=en&sa=X&ei=QmM4VZK6EpDAgwTt24CABw&ved=0CE8Q6AEwBw#v=onepage&q=Gompertz%20fish&f=false.]

Schnute, J. 1981. A versatile growth model with statistically stable parameters. Canadian Journal of Fisheries and Aquatic Sciences, 38:1128-1140.

Somers, I. F. 1988. On a seasonally oscillating growth function. Fishbyte 6(1):8-11. [Was (is?) from http://www.worldfishcenter.org/Naga/na_2914.pdf.]

Tjorve, E. and K. M. C. Tjorve. 2010. A unified approach to the Richards-model family for use in growth analyses: Why we need only two model forms. Journal of Theoretical Biology 267:417-425. [Was (is?) from https://www.researchgate.net/profile/Even_Tjorve/publication/46218377_A_unified_approach_to_the_Richards-model_family_for_use_in_growth_analyses_why_we_need_only_two_model_forms/links/54ba83b80cf29e0cb04bd24e.pdf.]

Troynikov, V. S., R. W. Day, and A. M. Leorke. Estimation of seasonal growth parameters using a stochastic Gompertz model for tagging data. Journal of Shellfish Research 17:833-838. [Was (is?) from https://www.researchgate.net/profile/Robert_Day2/publication/249340562_Estimation_of_seasonal_growth_parameters_using_a_stochastic_gompertz_model_for_tagging_data/links/54200fa30cf203f155c2a08a.pdf.]

Vaughan, D. S. and T. E. Helser. 1990. Status of the red drum stock of the Atlantic coast: Stock assessment report for 1989. NOAA Technical Memorandum NMFS-SEFC-263, 117 p. [Was (is?) from http://docs.lib.noaa.gov/noaa_documents/NMFS/SEFSC/TM_NMFS_SEFSC/NMFS_SEFSC_TM_263.pdf.]

Wang, Y.-G. 1998. An improved Fabens method for estimation of growth parameters in the von Bertalanffy model with individual asymptotes. Canadian Journal of Fisheries and Aquatic Sciences 55:397-400.

Weisberg, S., G.R. Spangler, and L. S. Richmond. 2010. Mixed effects models for fish growth. Canadian Journal of Fisheries And Aquatic Sciences 67:269-277.

Winsor, C.P. 1932. The Gompertz curve as a growth curve. Proceedings of the National Academy of Sciences. 18:1-8. [Was (is?) from http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1076153/pdf/pnas01729-0009.pdf.]

See Also

See Schnute for an implementation of the Schnute (1981) model.

Examples

Run this code
###########################################################
## Simple Examples -- Von B
( vb1 <- vbFuns() )
ages <- 0:20
plot(vb1(ages,Linf=20,K=0.3,t0=-0.2)~ages,type="b",pch=19)
( vb2 <- vbFuns("Francis") )
plot(vb2(ages,L1=10,L2=19,L3=20,t1=2,t3=18)~ages,type="b",pch=19)
( vb2c <- vbFuns("Francis",simple=TRUE) )   # compare to vb2

## Simple Examples -- Gompertz
( gomp1 <- GompertzFuns() )
plot(gomp1(ages,Linf=800,gi=0.5,ti=5)~ages,type="b",pch=19)
( gomp2 <- GompertzFuns("Ricker2") )
plot(gomp2(ages,L0=2,a=6,gi=0.5)~ages,type="b",pch=19)
( gomp2c <- GompertzFuns("Ricker2",simple=TRUE) )   # compare to gomp2
( gompT <- GompertzFuns("Troynikov1"))

## Simple Examples -- Richards
( rich1 <- RichardsFuns() )
plot(rich1(ages,Linf=800,k=0.5,a=1,b=6)~ages,type="b",pch=19)
( rich2 <- RichardsFuns(2) )
plot(rich2(ages,Linf=800,k=0.5,ti=3,b=6)~ages,type="b",pch=19)
( rich3 <- RichardsFuns(3) )
plot(rich3(ages,Linf=800,k=0.5,ti=3,b=0.15)~ages,type="b",pch=19)
( rich4 <- RichardsFuns(4) )
plot(rich4(ages,Linf=800,k=0.5,ti=3,b=0.95)~ages,type="b",pch=19)
lines(rich4(ages,Linf=800,k=0.5,ti=3,b=1.5)~ages,type="b",pch=19,col="blue")
( rich5 <- RichardsFuns(5) )
plot(rich5(ages,Linf=800,k=0.5,L0=50,b=1.5)~ages,type="b",pch=19)
( rich6 <- RichardsFuns(6) )
plot(rich6(ages,Linf=800,k=0.5,ti=3,Lninf=50,b=1.5)~ages,type="b",pch=19)
( rich2c <- RichardsFuns(2,simple=TRUE) ) # compare to rich2

## Simple Examples -- Logistic
( log1 <- logisticFuns() )
plot(log1(ages,Linf=800,gninf=0.5,ti=5)~ages,type="b",pch=19)
( log2 <- logisticFuns("CJ2") )
plot(log2(ages,Linf=800,gninf=0.5,a=10)~ages,type="b",pch=19)
( log2c <- logisticFuns("CJ2",simple=TRUE) ) # compare to log2
( log3 <- logisticFuns("Karkach") )
plot(log3(ages,L0=10,Linf=800,gninf=0.5)~ages,type="b",pch=19)
( log4 <- logisticFuns("Haddon") )


###########################################################
## Examples of fitting
##   After the last example a plot is constructed with three
##   or four lines on top of each other illustrating that the
##   parameterizations all produce the same fitted values.
##   However, observe the correlations in the summary() results.

## Von B
data(SpotVA1)
# Fitting the typical paramaterization of the von B function
fit1 <- nls(tl~vb1(age,Linf,K,t0),data=SpotVA1,start=vbStarts(tl~age,data=SpotVA1))
summary(fit1,correlation=TRUE)
plot(tl~age,data=SpotVA1,pch=19)
curve(vb1(x,Linf=coef(fit1)),from=0,to=5,col="red",lwd=10,add=TRUE)

# Fitting the Francis paramaterization of the von B function
fit2 <- nls(tl~vb2c(age,L1,L2,L3,t1=0,t3=5),data=SpotVA1,
            start=vbStarts(tl~age,data=SpotVA1,type="Francis",ages2use=c(0,5)))
summary(fit2,correlation=TRUE)
curve(vb2c(x,L1=coef(fit2)[1],L2=coef(fit2)[2],L3=coef(fit2)[3],t1=0,t3=5),
      from=0,to=5,col="blue",lwd=5,add=TRUE)

# Fitting the Schnute parameterization of the von B function
vb3 <- vbFuns("Schnute")
fit3 <- nls(tl~vb3(age,L1,L3,K,t1=0,t3=4),data=SpotVA1,
            start=vbStarts(tl~age,data=SpotVA1,type="Schnute",ages2use=c(0,4)))
summary(fit3,correlation=TRUE)
curve(vb3(x,L1=coef(fit3),t1=c(0,4)),from=0,to=5,col="green",lwd=2,add=TRUE)

## Gompertz
# Make some fake data using the original parameterization
gompO <- GompertzFuns("original")
# setup ages, sample sizes (general reduction in numbers with
# increasing age), and additive SD to model
t <- 1:15
n <- c(10,40,35,25,12,10,10,8,6,5,3,3,3,2,2)
sd <- 15
# expand ages
ages <- rep(t,n)
# get lengths from gompertz and a random error for individuals
lens <- gompO(ages,Linf=450,a=1,gi=0.3)+rnorm(length(ages),0,sd)
# put together as a data.frame
df <- data.frame(age=ages,len=round(lens,0))

# Fit first Ricker parameterization
fit1 <- nls(len~gomp1(age,Linf,gi,ti),data=df,start=list(Linf=500,gi=0.3,ti=3))
summary(fit1,correlation=TRUE)
plot(len~age,data=df,pch=19,col=rgb(0,0,0,1/5))
curve(gomp1(x,Linf=coef(fit1)),from=0,to=15,col="red",lwd=10,add=TRUE)

# Fit third Ricker parameterization
fit2 <- nls(len~gomp2(age,L0,a,gi),data=df,start=list(L0=30,a=3,gi=0.3))
summary(fit2,correlation=TRUE)
curve(gomp2(x,L0=coef(fit2)),from=0,to=15,col="blue",lwd=5,add=TRUE)

# Fit third Quinn and Deriso parameterization (using simple=TRUE model)
gomp3 <- GompertzFuns("QD3",simple=TRUE)
fit3 <- nls(len~gomp3(age,Linf,gi,t0),data=df,start=list(Linf=500,gi=0.3,t0=0))
summary(fit3,correlation=TRUE)
curve(gomp3(x,Linf=coef(fit3)[1],gi=coef(fit3)[2],t0=coef(fit3)[3]),
      from=0,to=15,col="green",lwd=2,add=TRUE)

## Richards
# Fit first Richards parameterization
fit1 <- nls(len~rich1(age,Linf,k,a,b),data=df,start=list(Linf=450,k=0.25,a=0.65,b=3))
summary(fit1,correlation=TRUE)
plot(len~age,data=df,pch=19,col=rgb(0,0,0,1/5))
curve(rich1(x,Linf=coef(fit1)),from=0,to=15,col="red",lwd=10,add=TRUE)

# Fit second Richards parameterization
fit2 <- nls(len~rich2(age,Linf,k,ti,b),data=df,start=list(Linf=450,k=0.25,ti=3,b=3))
summary(fit2,correlation=TRUE)
curve(rich2(x,Linf=coef(fit2)),from=0,to=15,col="blue",lwd=7,add=TRUE)

# Fit third Richards parameterization
fit3 <- nls(len~rich3(age,Linf,k,ti,b),data=df,start=list(Linf=450,k=0.25,ti=3,b=-0.3))
summary(fit3,correlation=TRUE)
curve(rich3(x,Linf=coef(fit3)),from=0,to=15,col="green",lwd=4,add=TRUE)

# Fit fourth Richards parameterization
fit4 <- nls(len~rich4(age,Linf,k,ti,b),data=df,start=list(Linf=450,k=0.25,ti=3,b=0.7))
summary(fit4,correlation=TRUE)
curve(rich4(x,Linf=coef(fit4)),from=0,to=15,col="black",lwd=1,add=TRUE)

## Logistic
# Fit first Campana-Jones parameterization
fit1 <- nls(len~log1(age,Linf,gninf,ti),data=df,start=list(Linf=450,gninf=0.45,ti=4))
summary(fit1,correlation=TRUE)
plot(len~age,data=df,pch=19,col=rgb(0,0,0,1/5))
curve(log1(x,Linf=coef(fit1)),from=0,to=15,col="red",lwd=10,add=TRUE)

# Fit second Campana-Jones parameterization
fit2 <- nls(len~log2(age,Linf,gninf,a),data=df,start=list(Linf=450,gninf=0.45,a=7))
summary(fit2,correlation=TRUE)
curve(log2(x,Linf=coef(fit2)),from=0,to=15,col="blue",lwd=5,add=TRUE)

# Fit Karkach parameterization (using simple=TRUE model)
log3 <- logisticFuns("Karkach",simple=TRUE)
fit3 <- nls(len~log3(age,Linf,L0,gninf),data=df,start=list(Linf=450,L0=30,gninf=0.45))
summary(fit3,correlation=TRUE)
curve(log3(x,Linf=coef(fit3)[1],L0=coef(fit3)[2],gninf=coef(fit3)[3]),
      from=0,to=15,col="green",lwd=2,add=TRUE)


#############################################################################
## Create expressions of the models
#############################################################################
# Typical von Bertalanffy ... Show as a stand-alone plot
growthFunShow("vonBertalanffy","Typical",plot=TRUE)
# Get and save the expression
( tmp <- growthFunShow("vonBertalanffy","Typical") )
# Use expression as title on a plot
lens <- vb1(ages,Linf=20,K=0.3,t0=-0.2)
plot(lens~ages,type="b",pch=19,main=tmp)
# Put expression in the main plot
text(10,5,tmp)
# Put multiple expressions on a plot
op <- par(mar=c(0.1,0.1,0.1,0.1))
plot(0,type="n",xlab="",ylab="",xlim=c(0,1),ylim=c(0,3),xaxt="n",yaxt="n")
text(0,2.5,"Original:",pos=4)
text(0.5,2.5,growthFunShow("vonBertalanffy","Original"))
text(0,1.5,"Typical:",pos=4)
text(0.5,1.5,growthFunShow("vonBertalanffy","Typical"))
text(0,0.5,"Francis:",pos=4)
text(0.5,0.5,growthFunShow("vonBertalanffy","Francis"))
par(op)

Run the code above in your browser using DataLab