|
|
Also, Carl Schwarz and George Seber recently published a comprehensive state-of-the-field review of current thinking on analysis of data from marked individuals. You can access it as an Adobe PDF file by click here. It is highly recommended reading, especially for those of you who haven't kept pace with the breathtaking rate of change in this area.
All of these programs are DOS/Windows based, although many have been ported to UNIX-based OS. None have been ported to the Mac (that I know of), so don't bother asking!
Last updated on July 25, 2001.
Program NOREMARK, developed by Gary White and
colleagues, is an interactive
program to estimate population size with marked animals. Resightings or
recaptures of marked animals, and the number of unmarked
animals observed on each occasion are used to compute
population estimates. Four different estimators are
provided in the program, each with different assumptions
about individual heterogeneity and immigration and
emigration to and from the study area.
Program ESTIMATE, developed by Gary White, was
one of the first programs developed to analyze recovery data. More
flexible than earlier programs (like BROWNIE), but nowhere near as
flexible as MARK (see below).
Program CAPTURE, developed by a group of people at
Colorado State, is the standard program for estimating
abundance of closed populations. It has a very good suite of
estimators, and is very flexible.
A program developed by Jim Hines and John Sauer for
post-hoc comparison of
survival/recapture/recovery rates for which variances and covariances are
available (analogous to means comparisoons in ANOVA). Uses the general
procedures described in Sauer & Williams (1989).
Program JOLLY (and its companion program JOLLYAGE - see
below) was developed by Jim Hines, and was one of the first programs to
allow estimation of both
survival/recapture rates, and population size, for a specified set of
models. Still in wide use since it was one of the first. However, no
flexibility in model testing.
Similar to program JOLLY (above) except that this
program handles 2 age-class models.
Program SURVIV, developed by Gary White, is
(essentially) a language - modeled
loosely on the SAS paradigm. You write code, call procedures, and
specifiy your model structure in a batch-oriented way. SURVIV is perhaps
infinitely flexible, and is used heavily as a research tool. However, you
need a FORTRAN compiler to run SURVIV, and a healthy appetite for
programming to get things to work. Not recommended for novices.
Derived by Jim Hines from Gary White's original SURVIV
code. The "MS" stands for "multi-state" or "multi-strata" - MS-SURVIV is
a program which allows you to reasonably easily estimate the
pobabilities of movement, simultaneously with survival and recapture.
Much more "user-friendly" than SURVIV, although still not overly easy
for the newcomer.
Program RELEASE, written by Gary White and colleagues at
Colorada State, was originally developed to estimate survival/resight for
a LARGE suite of fish mark-release experiments. However, it is also an
excellent program for simulation work, and provides a useful set of GOF
test diagnostic tools. Batch-oriented - easy to use, although the
documentation is hard to track-down (but see Chapter 5 in the
'MARK book' by Cooch & White.
The next generation...
These next 5 programs represent the current state-of-the-art. They are
the programs under the most active R&D, and incorporate many of the
newest ideas in the field. If you're going to spend time learning how to
use a particular program, any one of these 4 is where I would suggest
starting.
POPAN, written by Neil Arnason, Len Baniuk, and Carl
Schwarz stands for POPulation ANalysis. It
is a computer system for creating and maintaining files of mark-recapture
data from animal sampling experiments of open populations. Such files can
then be listed, or statistics can be
extracted from the entire file or subsets of the file, and a number of
standard analyses can be carried out, using
models of the Jolly-Seber type. It is closely modeled along the large,
batch-oriented statistical packages like SAS and BMDP. The only one of
this group that also estimates population abundance.
A new and unique program for analyzing CMR data -
developed by John Skalski and colleagues at the University of Washington.
Most notable for being
the first truly Windows-based program (both MS-Windoze and X11), and
(most especially) for using a proportional hazards scheme allowing use of
individual covariates. Only MARK (see below) can also do this. Major
limitation to SURPH is that it can't handle age-structured models (sensu
Pollock's models).
The grandfather of this group - the first program to
allows easy implementation of linear models to CMR analysis. Developed
by Jean Clobert, and more recently by Roger Pradel & Jean-Dominique
Lebreton in France. Very fast,
flexible, and easy to use. Much of the structure of some of the other
programs in this group (especially MARK and SURPH) can be clearly traced
back to SURGE. The newest version does automatic parameter counting
and calculation of the AIC.
The newest, and potentially the most complete, of the
packages. Developed by programming sensei Gary White at
Colorado State. Handles almost all kinds of analyses: both recovery and
recpature analysis (including open and closed-popualtion models),
telemetry analysis, multistate (i.e., movement models), and a variety of
other permutations on the standard paradigm (including the ability to
handle joint estimation from combined sources of data, and individual
covariates!). MARK does all this in a very intuitive Windows-based
application (intuitive if you have some experience with one of the other
"big" programs - most notably SURGE or SURPH).
To some degree, still a
work in progress, but pretty stable at present.
Requires Windows 10 or better, and a
pretty
hot-rod machine to run (I wouldn't try it on anything less than a
decent quad-core machine
with at least 8 Gb RAM).
All the technical documentation is in the help files - an introductory user's
manual
is in
prep (Cooch & White). Other online information about MARK is also available.
* maintained by G. White
Program Distance was written to analyze distance sampling data as
described
by Buckland et al.
(1993). This includes line transects, point transects (variable
circular
plots), trapping webs, and other related methods. The program now featues
an
intuitive windows-based interface, and a new version is in development
that
will include a built in GIS for automated survey design and spatial
modelling
of abundance, double platform methods (mark-recapture distance sampling)
and
several other features.
NOREMARK
(abundance estimation)
ESTIMATE
(recovery
analysis)
CAPTURE
(abundance estimation - closed populations)
CONTRAST
(comparison of survival rates)
JOLLY
(mark-recapute - open population - abundance & survival estimation)
JOLLYAGE
(open
population - mark-recapture - abundance & survival estimation with age
structure)
SURVIV
(programming language for survival estimation - open populations)
MS-SURVIV
(mark-recapture - movement models)
RELEASE
(mark-recapture - GOF testing - open populations)
MARK
[ Main MARK
Page*]
[ Local MARK Page]
DISTANCE (abundance
estimation from transect surveys)