PROGRAM INBRED: MODEL STRUCTURE, ASSUMPTIONS AND SENSITIVITY

• Main Authors: MacCracken, James G., Ash, Daniel E., Peek, James M., Van Ballenberghe, Victor
• Format: Article in Journal/Newspaper
• Language: English
• Published: Lakehead University 1996
• Subjects: Alces alces
• Online Access: http://alcesjournal.org/index.php/alces/article/view/859

We developed a stochastic, individual-based computer model that simulates the dynamics of an introduced moose (Alces alces) population, tracks pedigrees, and estimated genetic variables. The program language is Borland C++ and was compiled to run on IBM DOS compatibles with a 386 processor and 640 kilo-bytes (K) of random access memory (RAM). However, running time and RAM requirements increase exponentially with population size and most applications will need 16-32 mega-bytes (100K) of RAM, and a 486 or Pentium processor is recommended. Each time-step is one reproductive cycle, time-steps can be reiterated, and statistics are output for each time-step. Vital rates are fixed or density-dependent. A default file is used to specify vital rates, density-dependent fluctuations, bring behavior variables, inbreeding depression, number of iterations, and age- and sex-specific harvest rates. A data file defines the number of time-steps, age-sex harvest levels, and the characteristics of the base population. The major assumption of this model is the random mating of all sexually mature individuals. However, this assumption can be constrained by a number of breeding behavior variables. An inbreeding coefficient (Fp) is estimated for each individual and fetal death can occur based on the lethal equivalent function of Mills and Smouse (1994). A 2 allele, dominant-recessive function can be used to estimate founder effects. Model outputs include population size and Fd estimates by sex- and age-class, an index of effective population size (Ne) by sex, finite rate of increase (λ), percent twins, pregnancy rate, gene frequencies, and males/female. Sensitivity analyses indicated that variation in output was minimized with 400-700 iterations, density-dependent functions produced expected results, and that population and genetic variables were influenced most by calf and adult female mortality, adult female group size, calf sex ratio, and age to sexual maturity. The effects of translocations and harvesting moose on the adult sex ...