Increasing numbers of killer whale individuals use fisheries as feeding opportunities within subantarctic populations

Jolly-Seber capture-mark-recapture models to estimate Crozet regular and type-D Killer whales populations.

All codes are available for the simulation, the real data and the trends analysis

(1) Simulation code (codesimul.R) is running alone.

(2) Real data code (code_vraidata_revfin.R) makes use of three matrices that change depending on the ecotype:

- The Capture History (CH_2106_m.csv, CH_2106_d.csv), rows correspond to individuals and columns to years. For regular killer whales this matrix contains 3 possible observations (observed from the coast, observed depredating, non observed) and 2 observations for Type-D killer whales (observed depredating, non observed)

- The Marking status (Mk_2106_m.csv, Mk_2106_d.csv), rows correspond to individuals and columns to years. For both ecotype they contain three possible marking status as described in the ESM.

- The maturity status (St_2106_m.csv, St_2106_d.csv), rows correspond to individuals and columns to years. For both ecotype they contain four different states (non identify, new-born, juvenile and adult).

Two other matrices are the same for both ecotypes

- The standardized effort (Effort_stand_boat.csv) contains the yearly effort divided by the maximum number of photographs taken during the study period.

- The state matrix (z_init_cro2705.csv) is used to initiate the state matrix from the models.

(3) The trend_test.R file presents the extraction of model outputs (regular killer whales: CrozetFinrevfinfin.Rdata, Type-D killer-whales: DFinrevfinfin.Rdata) and their analysis trough different linear and non linear models.