dietOverlap: Computes indices of diet overlap between two species.

In droglenc/FSAmisc: Miscellaneous functions for simple fisheries stock assessment methods

Computes indices of diet overlap between two species.

Description

Computes various diet overlap indices between two species from summarized (i.e., not individual) diet data.

Usage

dietOverlap(
  diet1,
  diet2 = NULL,
  type = c("Horn", "Levins", "Morisita", "Pianka", "Schoener"),
  prey = NULL,
  num1 = NULL,
  num2 = NULL,
  N1 = NULL,
  N2 = NULL
)

## S3 method for class 'dietOverlap'
print(x, ...)

## S3 method for class 'dietOverlap'
summary(object, verbose = TRUE, digits = getOption("digits"), ...)

Arguments

diet1	A numerical vector of ‘amount’ (count or biomass) of prey items for the first predator. Items should be in the same order as amounts in diet2 and categories in prey.
diet2	A numerical vector of ‘amount’ (count or biomass) of prey items for the second predator. Items should be in the same order as amounts in diet2 and categories in prey.
type	A single string that indicates the type of diet overlap index to compute. See details.
prey	An optional string vector that contains the prey/diet category names. If supplied the prop matrix in the returned list will use these prey categories as row names. Items should be in the same order as items in diet1 and diet2.
num1	A numerical vector of the number of individuals of the first predator that consumed each prey item. Items should be in same order as amounts in diet1 and categories in prey.
num2	A numerical vector of the number of individuals of the second predator that consumed each prey item. Items should be in same order as amounts in diet1 and categories in prey.
N1	A single numeric value that is the total number of the first predator sampled.
N2	A single numeric value that is the total number of the second predator sampled.
x	A dietOverlap object.
...	Additional arguments for the S3 methods. Not implemented for summary.
object	A dietOverlap object.
verbose	A single logical that indicates whether more verbose summary information should be printed.
digits	A single numeric that indicates the number of digits to which the results should be printed.

Details

NEED MORE DETAILS HERE

Value

The main function returns a list with the following three items:

• type A single string that indicates the type of diet overlap indice used.

• doi A single numeric with the diet overlap index value is returned for all types except for ‘Levins’ where a numeric vector of length two is returned with the overlap of the first predator on the second as the first value and the overlap of the second predator on the first as the second value.

• propdiet A matrix that contains the proportion of the total diet in each of the diet items for each predator.

Note

Only the Pianka measurements have been tested against results from other softwares.

Author(s)

Derek H. Ogle, [e-mail]

References

Horn, H.S. 1966. Measurement of overlap in comparative ecological studies. American Naturalist 100:419-424.

Krebs, C.J. 1989. Ecological Methodology. Harper Collins, New York.

Levins, R. 1968. Evolution in changing environments: Some theoretical explorations. Princeton University Press, Princeton.

Morisita, M. 1959. Measuring of interspecific association and similarity between communities. Memoirs of the Faculty of Science of Kyusha University. Series in Evolutionary Biology 3:65-80.

Pianka R.D. 1973. The structure of lizard communities. Annual Review of Ecology and Systematics 4:53-74. Schoener, T.W. 1970. Nonsynchronous spatial overlap of lizards in patchy habitats. Ecology 51:408-418.

Smith, E.P. and T.M. Zaret. 1982. Bias in estimating niche overlap. Ecology 63:1248-1253.

See Also

See piankabio in pgirmess for similar functionality.

Examples

# Hypothetical data -- prey categories and biomasses in diets
names <- c("bluegill","perch","minnows","bullheads","insects","zooplankton")
lmb <- c(55,35,23,7,3,1)
wae <- c(23,45,2,17,7,2)

# demonstrate different indices (see below for Morisita)
dietOverlap(lmb,wae,prey=names,type="Horn")
dietOverlap(lmb,wae,prey=names,type="Levin")
dietOverlap(lmb,wae,prey=names,type="Pianka")
dietOverlap(lmb,wae,prey=names,type="Schoener")

# demonstrate summary()
do1 <- dietOverlap(lmb,wae,names,type="Horn")
summary(do1)
summary(do1,digits=3)
do2 <- dietOverlap(lmb,wae,names,type="Levin")
summary(do2)

# demonstrate using a single matrix rather than two separate vectors
diet <- cbind(lmb,wae)
rownames(diet) <- names
dietOverlap(diet,prey=rownames(diet),type="Horn")

# Continuation of first example with 'extra' info required for Morisita's index
# Hypothetical numbers of fish with each food item (occurrence)
num.lmb <- c(17,13,4,7,24,13)
num.wae <- c(37,12,23,8,9,13)
# Hypothetical total numbers of each predator
N.lmb <- 30
N.wae <- 40

dietOverlap(lmb,wae,prey=names,type="Morisita",num1=num.lmb,num2=num.wae,N1=N.lmb,N2=N.wae)

## An extended example which requires summarization of raw data
data(TroutDietSL)
# add percentage per fish, add zeroes for empty stomachs
TroutDietSL$pvol.mysis <- TroutDietSL$vol.mysis/TroutDietSL$vol.ttl
TroutDietSL$pvol.oi <- TroutDietSL$vol.oi/TroutDietSL$vol.ttl
TroutDietSL$pvol.fish <- TroutDietSL$vol.fish/TroutDietSL$vol.ttl
# add empty variable
TroutDietSL$empty <- ifelse(TroutDietSL$vol.ttl==0,"YES","NO")
# create TL from SL for Bull Trout using formula in paper -- TL=4.403+1.118SL
TroutDietSL$tl[TroutDietSL$species=="BLT"] <- 4.403+1.118*TroutDietSL$sl[TroutDietSL$species=="BLT"]
# add LCat length categories
TroutDietSL$LCat <- FSA::lencat(TroutDietSL$tl,breaks=c(0,350,500,650,950),right=TRUE)

# isolate the non-empty fish (as the authors did)
TroutDietSL1 <- subset(TroutDietSL,empty=="NO")
with(TroutDietSL1,table(species,LCat))

# summarize by computing the mean percent of the three diet items
mn.mysis <- aggregate(pvol.mysis~species*LCat,data=TroutDietSL1,FUN=mean)
mn.oi <- aggregate(pvol.oi~species*LCat,data=TroutDietSL1,FUN=mean)
mn.fish <- aggregate(pvol.fish~species*LCat,data=TroutDietSL1,FUN=mean)
mndiet <- cbind(mn.mysis,mn.oi[,"pvol.oi"],mn.fish[,"pvol.fish"])
colnames(mndiet)[3:5] <- c("mysis","oi","fish")
# reorganize the result
mndiet <- reshape(mndiet,direction="long",varying=c("mysis","oi","fish"),v.names="mnprop",
                  idvar=c("species","LCat"),timevar="item",times=c("mysis","oi","fish"))
mndiet <- reshape(mndiet,direct="wide",v.names="mnprop",idvar=c("item","LCat"),timevar="species")

# Now compute diet overlap between same sized fishes of the two species
dietOverlap(mndiet[mndiet$LCat=="0",3:4],prey=levels(mndiet$item),type="Schoener")
dietOverlap(mndiet[mndiet$LCat=="350",3:4],prey=levels(mndiet$item),type="Schoener")
dietOverlap(mndiet[mndiet$LCat=="500",3:4],prey=levels(mndiet$item),type="Schoener")
dietOverlap(mndiet[mndiet$LCat=="650",3:4],prey=levels(mndiet$item),type="Schoener")

droglenc/FSAmisc documentation built on Jan. 8, 2023, 12:59 a.m.