Kendall {popbio} | R Documentation |
This function finds the best estimates of mean and environmental variance for beta-binomial vital rates, using a brute force search for the best adjusted estimates from a very large number of combinations of different possible mean and variance values.
Kendall(rates, grades=1000, maxvar=0.2,minvar=0.00001, maxmean=1, minmean=0.01)
rates |
a matrix or dataframe with four columns: Rate identifier, Year, Total number of starting individuals, Number growing (or surviving). |
grades |
number of different levels of means and variances to try, default is 1000 |
maxvar |
maximum variance to search over, default is 0.20. The maximum ever possible is 0.25 and searching a narrower range will improve the accuracy of the answer. |
minvar |
minimum variance to search, default is 0.00001. |
maxmean |
maximum limit on the mean values to search, default 1 |
minmean |
minimum limit on the mean values to search, default 0.01 |
A list with estimates and confidence intervals
est |
a matrix with 5 columns: (1) estimated mean, (2) Kendall's MLE mean, (3) estimated variance, (4) Kendall's MLE variance, (5) Kendall's unbiased MLE variance. |
ci |
a matrix with 95% confidence limits for the Kendall's mean and unbiased variance estimates with 4 columns: (1) low and (3) high mean limits, (3) low and (4) high variance limits. |
Note that it may deliver warning messages of : 'no finite arguments to min; returning Inf', indicating use of very low values for variance, but this is not a malfunction.
Adapted to R from Morris & Doak (2002: 267-270) by Patrick Nantel.
converted Matlab code from Box 8.2 in Morris and Doak (2002)
Kendall, B. E. 1998. Estimating the magnitude of environmental stochasticity in survivorship data. Ecological Applications 8(1): 184-193.
Morris, W. F., and D. F. Doak. 2002. Quantitative conservation biology: Theory and practice of population viability analysis. Sinauer, Sunderland, Massachusetts, USA.
## desert tortoise input from Box 8.2 - compare results to Table 8.3 tor<-data.frame(rate=rep(c("g4","g5","g6"),each=3), year=rep(1:3,3), ## representing 70s, early 80s, late 80s start=c(17,15,7,22,19,4,32,31,10), grow=c(8,1,0,5,5,0,2,1,0)) ## use fewer grades for faster loop tor.est<-Kendall(tor, grades=200) tor.est data(woodpecker) wp.est <- Kendall(woodpecker, grades=200) wp.est