Species-Presence exercise 4 - Multi-Season data

This exercise is designed to show how to run program PRESENCE to compute species presence, detectability, colonization and extinction estimates from 'presence-absence' data.

Input data consists of 'detection-histories' of individuals at survey sites.

Running the program

Start program PRESENCE.

Click 'File/New Project' to start a new analysis. A dialog box will appear asking for a title, filename, number of sites,etc.

Now, start your spreadsheet program, and load the sample data file named 'NSO_pg209.xls'. Determine the number of sites, years, and total number of surveys in the file. Select all data cells (using the mouse), and copy to the clipboard ('Edit'/'Copy').

Next, return to program PRESENCE and enter the number of sites, total number of surveys, and the number of surveys per year in the appropriate boxes. Then, click 'Data Input Form' and click in the first cell of the input data form (site 1, 1-1). Go to the 'Edit' menu, and click 'Paste', 'paste values'.

Save the input by clikcing 'File/Save as' and name the file 'NSO_pg209.pao'. Close the Input Data Form window by clicking the X in the upper right corner or clicking 'File/Close'.

Next, click the 'Click to select file' button on the 'Enter Specifications' window. When the dialog box appears, select the file just created (NSO_pg209.pao) and click 'Open'. Click in the 'Title' box and enter 'Multi-season example from page 209 Northern Spotted Owls'. Click 'OK'.

You're now presented with a 'Results Browser' window where a summary of each model will be saved. To run our first model, click 'Run/Analysis:multi-season' from the menu-bar.

When the 'Setup Numerical Estimation Run' window appears, click the design-matrix window to bring it to the front. The first model to run will be one where colonization and extinction are constant over time and detection probabilities are year-specific. With the default parameterization, there is only an initial occupancy parameter (psi), and colonizatin (gam) and extinction (eps) are preset to constant over time, so we only need to make detection year-specific. Click the 'Detection' tab, go to the 'Init' menu, and click 'Seasonal Effects'. Then, go back to the 'Setup Numerical Estimation Run' window, enter 'psi,gamma(.),eps(.),p(yr)' for the model name and click 'OK to Run'.

Next, run a model which is the same as the last one except that colonization (gamma) and extinction (eps) are year-specific. Click 'Run/Analysis:Multi-season', make sure the first model parameterization option is selected and click on the design-matrix window. We can retrieve the design-matrix from the previous model by clicking 'Retrive model' and select the previous model from the list. Next, click the 'Colonization' tab, click 'Init/Full Identity' to make colonization different for each year. Do the same for extinction. Go back to the 'Setup Numerical Estimation Run' window and enter 'psi,gamma(yr),eps(yr),p(yr)' for the model-name and click 'OK to Run'.

Now, we'll run the same model with a different parameterization. This time, click 'Run/Analysis:multi-season', and select the 2nd model parameterization (seasonal occupancy, colonization and detection). In the design-matrix window, click the 'occupancy' tab, and click 'Init/Full Identity' to make occupancy year-specific. Do the same for colonization (gamma), and make detection year-specific by clicking the detection tab and clicking 'Init/seasonal effects'. Then, change the model-name to 'psi(yr),gam(yr),p(yr)' and click 'OK to Run'.

The results of this model should match the previous model. Looking at the output, you can see that the model did not produce reasonable estimates and standard errors. For the combination of this particular model and data, the program needs a little help. Repeat the steps to run this model again, except check off the box allowing you to input initial parameter values. Enter '.6,.6,.6,.6,.6,.1,.1,.1,.1,.1,.1,.1,.1' for the initial values and run.

Next, let's run a model where occupancy is constant, colonization is constant, and detection is year-specific. (Note that we're talking about seasonal occupancy being constant, not the initial occupancy parameter.) For this, we need to select the 2nd model parameterization option (Seasonal occupancy, colonization and detection). Click 'Run/Analysis:multi-season' and select the 2nd model parameterization option. When you do this, the 'extinction' tab will disappear from the design-matrix window, and rows will be added to the occupancy table to allow occupancy to (potentially) differ each year. By default, there is only one column in the occupancy table, so occupancy will be constant over time. If you click the 'Colonization' tab, you'll see that colonization is constant over time by default. Click the Detection tab and click 'Init/Seasonal effects' to make detection year-specific. Go back to the 'Setup Numerical Estimation Run' window, change the model name to 'psi(.),gamma(.),p(yr)' and click 'OK to Run'.