Generalized Additive Mixed Model (GAMM) analysis results for green turtle catch data from the Caribbean coast of Nicaragua.

For each model its AIC value and the difference in AIC values between the top ranked model (value in bold) and other models (ΔAIC) is shown. Models are sorted by analysis type: overall, regional (when applicable), and community level (with the “best” model in terms of smallest AIC value in bold italics by analysis level and in bold underline overall).
Results are shown for the three data sets analyzed: COM  =  principal fishing community data, AW COMCTR  =  commercial center data for the Awastara community, PCFA  =  Pearl Cays fishing area data. Explanatory variable acronyms are: TM  =  TIME, MN  =  MONTH, RG  =  REGION, CM  =  single or aggregated community, TIME by  =  interaction between TIME and either REGION or community.
For terms included in a model, p-values were indicated as follows: *** = <0.001 or ** = <0.01 and values >0.1 are designated as non-significant (NS), with results for the three data sets considered separated by a semi-colon.
a“X” indicates when temporal autocorrelation was treated using an autoregressive model of order one.