There has been a considerable amount of recent work on software to perform population analysis, particularly in terms of estimation of abundance, and both survival and recruitment rates (using both capture-recapture and recovery models). The following is a fairly complete list of available programs and related information. Please let me know of any useful additions. I also recommend taking a look at the Patuxent Software Archive Page. In addition, there is an online discussion forum which concerns theory and application of many of these programs.

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.


NOREMARK (abundance estimation)

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.

ESTIMATE (recovery analysis)

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).

CAPTURE (abundance estimation - closed populations)

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.

CONTRAST (comparison of survival rates)

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).

JOLLY (mark-recapute - open population - abundance & survival estimation)

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.

JOLLYAGE (open population - mark-recapture - abundance & survival estimation with age structure)

Similar to program JOLLY (above) except that this program handles 2 age-class models.

SURVIV (programming language for survival estimation - open populations)

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.

MS-SURVIV (mark-recapture - movement models)

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.

RELEASE (mark-recapture - GOF testing - open populations)

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

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.

SURPH

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).

SURGE

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.

MARK [ Main MARK Page*] [ Local MARK Page]

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 Vista/7 or better, and a pretty hot-rod machine to run (I wouldn't try it on anything less than a decent dual-core machine with at least 2 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

DISTANCE (abundance estimation from transect surveys)

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.