Encounter Histories Format


Encounter Histories Format

Each data type of Program MARK requires a primary form of data entry provided by the encounter histories.  Encounter histories can be either information on only live encounters (LLLL) or information on both live and dead encounters (LDLDLDLD).  In addition, some data types allow a summary format that reduces the amount of input.  The following table summaries the input for each data type as specified in the encounter histories file.

Data TypeEncounter History Format Additional Input Format, Notes
Recaptures only (CJS) LLLL allows . in enc. hist. 
Recoveries only LDLDLDLD Recovery matrix 
Both (Burnham) LDLDLDLD  
Known Fate LDLDLDLD Known fate 
Closed Captures LLLL  
BTO Ring Recoveries LDLDLDLD Recovery matrix 
Robust Design LLLL  
Both (Barker) LDLDLDLD  
Multi-strata Recaptures  LLLL  
Brownie et al. Recoveries LDLDLDLD Recovery matrix 
Jolly-Seber LLLL  
Huggins’ Closed Captures LLLL  
Robust Design (Huggins Est.) LLLL  
Pradel Recruitment only LLLL  
Pradel Survival and Seniority LLLL  
Pradel Survival and Lambda LLLL  
Pradel Survival and RecruitmentLLLL   
Barker Robust Design LDLDLDLD  
VPA — Virtual Population Analysis LDLDLDLD  
Multi-strata — Live and Dead Enc.LDLDLDLD   
Closed Captures with Het. LLLL  
Full Closed Captures with Het. LLLL  
Nest Survival LDLDLDLD Nest survival  
Occupancy Estimation LLLL allows . in enc. hist. 
Robust Design Occupancy EstimationLLLL allows . in enc. hist.  
Open Robust Design Multi-strataLLLL   
Closed Robust Design Multi-strataLLLL   
Density using radio-tracking LLLL grid total allows . in enc. hist. 

Following each encounter encounter history is the number of animals with this encounter history for each group (known as frequencies).  One or more spaces are used to separate the encounter history from the frequencies. Note that frequencies are counts, and are not limited to just the values of 0 and 1. In other words, the frequencies are not indicator or dummy variables to specify the group identification.  The frequencies are used in the log likelihood to provide the observed number of animals with this cell probability in the form observed*log(probability).

If individual covariates are included, these values follow the frequencies for each group.  Typically with individual covariates, only one frequency will be non-zero, usually with a value of 1 because only one animal would have the specific set of individual covariate values.  One or more speces are used to separate the individual covariates.

A semicolon terminates the input for each encounter history.