Table 1 Criteria for evaluating performance among model estimates for regional occupancy
Test | Description | Interpretation |
|---|---|---|
Mean weighted occupancy error (MOE) | Weighted mean of absolute difference between estimated occupancy for a region and the corresponding actual occupancy averaged across regions. Error for each region was weighted by the proportional use of each region | Larger error implies a less accurate occupancy model |
Regional occupancy error (ROE) | Difference between estimated occupancy and actual occupancy for a given region | Positive values imply overestimates of regional occupancy; negative values imply underestimates of regional occupancy |
Absolute regional occupancy error (aROE) | Absolute difference between estimated occupancy and actual occupancy for a given region | Larger error implies a less accurate regional occupancy model |
Daily region assignment error | Percent incorrect model-based daily regional assignment for each simulation. When a simulated track or model indicated occupancy of more than one region on a given day, the region used most for each day was used for the comparison | Larger values imply model is not accurately estimating daily location of individual fish |
Within-region occurrence error | Difference between the estimated number of distinct simulated tracks that occurred within a region based on movement model predictions and actual positions of all tracks | Positive values imply overestimates of distinct simulated tracks; negative values imply underestimates of distinct simulated tracks |