Fig 1.
Distribution of the longline fishing sets conducted during the first (2009–2011) and second (2012–2016) phase.
The 200 m (solid line) and the 3000 m (dashed line) isobaths are represented.
Fig 2.
Image from the stern of a fishing vessel of the Uruguayan tori line deployed in situ showing the long and short streamers and the position to starboard in relation to the centrally deployed fishing mainline.
Table 1.
Number of birds incidentally captured in each treatment during Phases 1 and 2.
The observed fishing effort was as follows: Phase 1: with tori line = 51 sets and 52,371 hooks, without a tori line = 49 sets and 50,613 hooks; Phase 2: with tori line = 73 sets and 47,554 hooks.
Table 2.
Estimated coefficients and standard errors (SE) of the GLMs (Negative Binomial) for the bycatch rate of seabirds.
The Akaike information criteria (AIC) values and the dispersion parameters are shown for each model. All models were fitted using the log link function. Note that for every factor, one category is fixed (intercept), which serves as the standard for comparisons with other levels.
Table 3.
Estimated coefficients and standard errors (SE) of the GLMM (Binomial) for the ruptures of the tori line.
The rate of change in the odds and 95% confidence limits (c.l.) is presented for each variable. Variance of random effect “fishing trip” was 0.33 and standard deviation was 0.57.
Fig 3.
Mean aerial coverage of the tori line in phases 1 and 2.
Error bars represent ± 1 standard deviation (SD). The number above each bar indicates the sample size.
Fig 4.
Bird capture per unit of effort (BCPUE; birds/1000 hooks) during day and night longline sets without tori line (No TL) and with tori line (TL).
The number above each bar indicates the sample size.