The Heisey and Fuller (1985) method is a generalization of the basic approach of Mayfield (1961) and Trent and Rongstad (1974) to determine unbiased estimates of cause-specific mortality rates. Heisey and Fuller (1985) also developed a computer program MICROMORT to perform the calculations.
The Heisey and Fuller (1985) method can be described by the following equations where the subscript i (i = 1, 2, ..., I) indicates the ith interval being considered. The maximum likelihood estimate for survival in the ith interval (ŝi) is given by the following equation where xi is the total number of transmitter days (number of collar animals multiplied by the number of days) and yi is the total number of deaths
$$\hat{s}_i = \frac{x_i - y_i}{x_i}$$
Where Li is the length of days in the interval then the estimated survival rate (Ŝi) for the ith interval is
Ŝi = ŝiLi The total survival rate, S*, over all, I, intervals is simply the product of the Ŝi’s
$$ \hat{S}^{*} = \prod_{i=1}^{I}{\hat{S}_{i}} $$
The standard error is calculated using the Taylor series approximation method. The sampling distribution of Ŝ* is quite skewed therefore the 95% Wald confidence interval (Millar 2011) is calculated on the log scale (before being back-transformed)
log (Ŝ*) ± 1.96 Ŝ* − 1 SE(Ŝ*)
The nonparametric staggered entry Kaplan-Meier, which uses discrete time steps, is defined algebraically below (Pollock et al. 1989). The probability of surviving in the ith timestep is given by the equation
$$ \hat{S}_i = 1 - \frac{d_i}{r_i} $$
where di is the number of mortalities during the period and ri is the number of collared individuals at the start of period. The estimated survival for any arbitrary time period t is given by
Ŝ(t) = ∏i = 1(Ŝi, t) where Ŝi, t is the survival during the ith timestep of the tth period.
One method to estimate the standard error (Cox and Oakes 1984) uses Greenwood’s formula
$$ \text{SE}(\hat{S}(t)) = \sqrt{[\hat{S}(t)]^2 \sum{\frac{d_{i}}{r_{i}(r_{i} - d_{i})}}} $$
However they also propose a simpler alternative estimate “which is better in the tails of the distribution” (Pollock et al. 1989)
$$ SE(\hat{S}(t)) = \sqrt{ \frac{[\hat{S}(t)]^2 - [1 - \hat{S}(t)]}{r(t)}} $$
Although Pollock et al. (1989) proposed calculating 95% Wald confidence intervals (Millar 2011) this formulation can result in lower confidence limits below zero or upper confidence limits above one when using the standard equation
Ŝ(t) ± 1.96 SE(Ŝ(t)) Confidence intervals were also computed using bootstrapping with 1,000 (McLoughlin et al. 2003) or 10,000 replicates (Hervieux et al. 2013).
To account for censoring DeCesare et al. (2012) estimated female survival with nonparametric methods and “then adjusted the input annual sample sizes of animals and survival events to produce equivalent mean and variance estimates using a binomial variance estimator (Morris and Doak 2002)” (DeCesare et al. 2012).
There are other less used methods like the Skalski staggered entry equation (Skalski, Ryding, and Millspaugh 2005) and the known fate binomial model in program MARK. The known fate model provides a similar estimate to the Kaplan-Meier estimator when it is constrained to estimate yearly survival from the product of monthly survival rates. The advantage of the known fate approach is it allows comparison of alternative models with, for example, constant survival vs yearly and/or monthly variation. As well as the Heisey and Fuller method, Edmonds (1988) also used the method described by Gasaway et al. (1983) to provide estimates and confidence intervals.
The mean ratio, R̂, can be calculated from, ȳ, the mean number of calves and, c̄, the mean number of cows (Cochran 1977; Thompson 1992; Krebs 1999)
$$ \hat{R} = \frac{\bar{y}}{\bar{c}} $$ Recruitment rates were adjusted by DeCesare et al. (2012) to provide an estimate of recruitment that assumes the calves surveyed in the March surveys have recruited to adults. DeCesare et al. (2012) argued uncorrected calf-cow ratios did not account for recruits at year end and therefore provide an overoptimistic estimate of recruitment and subsequent population trend. In the equation derived by DeCesare et al. (2012), the age ratio, X, is commonly estimated as the number of calves, nj, per adult female, naf, observed at the end of a measured year, such that
$$ X = \frac{n_{j}}{n_{af}} $$ and $$ \frac{X}{2} = \frac{n_{jf}}{n_{af}} $$
where X/2 estimates the number of female calves (njf) per adult female assuming a 50:50 sex ratio, proper adjustments of X/2 to a calves/(calves + adults) ratio is necessary according to
$$ R_{RM} = \frac{n_{jf}}{n_{f}} = \frac{ \frac{X}{2} }{1 + \frac{X}{2}} $$
Where nf = njf + naf, or the total number of females of all age classes, including calves, counted at the end of the measurement year.
The standard error of R as defined by the ratio (x/y) is given by
$$ SE(\hat{R}) = \frac{\sqrt{1-f}}{\sqrt{n}y} \sqrt{\frac{(1-f)(\sum{x^{2}-2\hat{R}} \sum{xy+\hat{R}^{2}\sum{y^{2}}})}{n-1}} $$
Where f is the sampling fraction n/N, n is the sample size, N is the sum of the sample sizes, and ȳ is the observed mean of Y measurement (denominator of ratio) (Cochran 1977; Thompson 1992; Krebs 1999).
Bootstrapping was used to generate confidence intervals using 1,000 (McLoughlin et al. 2003) and 10,000 replicates (Hervieux et al. 2013, 2014).
A binomial distribution was used by Rettie and Messier (1998).
The three main methods used for calculating, λ, population growth are the original recruitment-mortality R/M formula described by Hatter and Bergerud (1991) using raw calf cow ratios of R; the adjusted recruitment RRM formula of DeCesare et al. (2012) and a Lefkovich matrix model (Fryxell et al. 2020)
The original recruitment-mortality formula of Hatter and Bergerud (1991)
$$ \lambda = \frac{S}{1-R} $$
The adjusted recruitment rate of DeCesare et al. (2012) will provide a better estimate of λ then using uncorrected calf-cow ratios from spring surveys (as discussed in the recruitment section).
A Lefkovich matrix (Fryxell et al. 2020) which estimates lambda based on the maximum eigenvalue of a Lefkovich matrix of stage-based survival and birth rates. Where the matrix-based population viability analysis (PVA) model used population-wide estimate of annual survival of adult and yearling females (S) and successful offspring recruitment (S × B, where B is birth rate of female offspring) in each study area to fill the elements of a Lefkovich matrix (L) where Bj is offspring recruitment rate stemming from age class j, and Sj is the annual survival rate of age class j, where j = 0, 1 and a for new-born calves, yearlings, and adults (Fryxell et al. 2020).
$$ L = \left[\begin{array}{ccc} 0 & S_1 \times B_1 & S_a \times B_a \\ S_0 & 0 & 0 \\ 0 & S_1 & S_a \end{array} \right] $$
DeCesare et al. (2012) also considered stage-based matrix models and provides discussion on how matrix models relate to the standard S/(1 − R) equation.
Confidence intervals for population growth were calculated through simulation. One way was through Monte Carlo simulations by randomly drawing from the annual survival and recruitment distributions (Rettie and Messier 1998; Hervieux et al. 2013, 2014) using programs like Excel’s MATLAB or Poptools. Another method was through bootstrapping (McLoughlin et al. 2003).