Percentage deviance explained by predictors in GAM and DistLM analyses.

Percentage of deviance explained in marginal tests using Generalised Additive Models for the proportion of orange roughy containing prey, %S, fish weight during the survey in 2010 and surveys between 1986 and 2010, and for diet composition using the DistLM analysis. Approximate significance of predictors: NS, >0.05;*≤ 0.05;**≤ 0.01;***≤ 0.001.本数据集记录了两类分析中边际检验的偏差解释百分比：其一为采用广义可加模型（Generalised Additive Models），针对2010年调查及1986至2010年间历次调查中含猎物的橘棘鲷（orange roughy）占比（%S）、鱼体体重开展的分析；其二为采用距离线性模型分析（DistLM analysis）的食性组成分析。各预测变量的近似显著性标注规则如下：NS代表P>0.05；*代表P≤0.05；**代表P≤0.01；***代表P≤0.001。