4 Results
4.1 Results of trap study
During the whole 2018-2019 period, a total number of 28 trips have been undertaken to the Saba Bank to either deploy traps, haul them or both. In total 275 hauls were done and a total of 182 lionfish have been caught. Most effort has been done in 2018, as nearly twice as much trap were hauled during almost thrice as many fishing trips to the Saba Bank. In total 129 traditional traps have been hauled and 146 funnel traps have been hauled, regardless of FAD colour. A higher number of lionfish have been caught during 2018 (166) than in 2019 (16). Another interesting difference is the amount of trap which have been hauled empty. 7 out of 183 have been hauled empty in 2018 against 13 out of 92 in 2019. Table 5 gives a descriptive overview of the efforts and the catch rates of lionfish and bycatch during the research, with a distinction made between 2018 and 2019.
Table 5. Descriptive data and frequencies of the 2018 and 2019 data collection sessions
2018 2019 Total
Trap design and number of lionfish
No significant differences were found between the number of lionfish and trap design (P=0.105).
Although not significant, the mean number of lionfish in the funnel trap tended to be lower in the funnel trap than in the traditional trap design. The traditional trap had an average of 0,82 lionfish per trap while the funnel trap had an average of 0,52. The highest amount of lionfish caught in a single trap haul was nine, the trap design was traditional, and it soaked for 21 days. The second and third highest amounts were eight and seven and both hauls were done using the funnel trap design and soaked for seven days each. Out of the 275 total trap hauls 63 contained at least one lionfish. Only 6 of these traps were hauled in 2019 and the other 57 in 2018 (see Figure for a lionfish caught in a traditional design lionfish trap).
24 Trap design and number of bycatch and bycatch species
The Mann-Whitney U nonparametric test found a significant difference between the two trap designs and the number of bycatch (P = 0,000). Graph 2 shows the mean number of bycatch individuals per trap design. The traditional design had an average of 12,38 individuals of bycatch per trap haul. The funnel design had an average of 4,71 individuals of bycatch per trap haul.
Another significant difference was found between the two trap designs and the number of bycatch species (P=0,000). Graph 3 shows the difference between the mean number of bycatch species and the two different trap designs. The traditional trap design had an average of 3,98 bycatch species per trap. The funnel trap design had an average of 2,52 bycatch species per trap.
Graph 2. Bar chart showing the mean number of bycatch caught per trap design.
Graph 3. Bar chart showing the mean number of bycatch species per trap design
25 Depth categories and number of lionfish
The Kruskal-Wallis H nonparametric test found no significant differences between the depth categories and the number of lionfish (P=0,079). Although no significant difference was found, results show that the depth distribution of caught lionfish closely resembles the depth distribution of the trap hauls (see Graph 1 in methods).
Depth categories and number of bycatches and bycatch species
A significant difference was found between the depth categories and the number of bycatch (P = 0,000). Graph 4 shows the mean number of bycatch per depth category divided over trap design.
Another significant difference was found between the depth categories and the number of bycatch species (P=0,007). Graph 5 shows the mean number of bycatch species per depth category divided over trap design. Post-hoc analysis of the Kruskal-Wallis H test results revealed multpiple significant differences between many depth categories and number of bycatch individuals and species. An overview of all signifcant differences and associated P-values can be found in Appendix IV.
The bar-charts have put both bycatch variables per depth category and observed was a clear decline in both the number of bycatch individuals and the number of species as fishing depth increases. On average, most bycatch species and bycatch individuals have been caught at a depth of 121 to 150 feet.
A distinction between trap design was also made to emphasize the significance possible difference in catch rates on certain depths. What can be seen is that for the mean number of bycatch individuals decreases between 301 feet and 390 feet when using the funnel design, while it increases when using the traditional design (Graph 4). The exact same pattern can be seen when comparing the fluctuations of the mean amount of bycatch species per trap design on the same depth of 301-390 feet (Graph 5).
26
Graph 4. Clustered bar chart of mean number of bycatch per depth category divided over trap design..
Graph 5. Clustered bar chart for mean number of bycatch species per depth category divided over trap design.
27 Bycatch composition
During the research a wide variety of bycatch species have been caught. These range from smaller reef fishes like surgeon fishes, butterfly fishes, and cowfishes to larger species such as snappers, groupers and jacks. Table 6 and Graph 6 show an overview of the most common caught species of bycatch during the 2018-2019 study. As mentioned before only bycatch species who represented more than one percent of the dataset were taken into account. The total frequency per species and the percentage of the total amount were recorded. By far the most frequent bycatch species is the blackfin snapper (L. bucanella) with 1011 individuals (46,9% of total). The blackfin snapper is accompanied by other snapper species yellow-eye snapper (S. ruberrimus) with 249 individuals (11,6%
of total) and the vermillion snapper (R. aurorubens) with 231 individuals (10,7% of total). The red hind (E. auttatus) is the fourth most abundant bycatch species with 191 individuals (8,9% of total).
Interestingly these four species are also the main targets in the redfish fisheries of Saba and are all commercially interesting species and comprise 78,09% of total bycatch The longspine squirrelfish (H.
rufus) is also highly abundant with 163 individuals (7,6%), however, this species is not commercially interesting and is always discarded immediately by fishermen when caught in fish traps.
Species Latin name Frequency Percentage of total (%)
Banded Butterflyfish Chaetodon striatus 48 2,2
Blackfin Snapper Lutjanus buccanella 1011 46,9
Blue Tang Acanthurus Coeruleus 43 2,0
Coney Grouper Cephalopholis fulva 37 1,7
Red Hind Epinephelus guttatus 191 8,9
Spotted Goatfish Pseudupeneus maculatus 49 2,3 Longspine Squirrelfish Holocentrus rufus 163 7,6 Vermillion Snapper Rhomboplites aurorubens 231 10.7
White Grunt Haemulon plumierii 93 4,3
White-spotted Filefish Cantherhines macrocerus 39 1,8
Yellow-eye Snapper Lutjanus vivanus 249 11,6
Table 6. Frequency of bycatch species caught in traps.
28
A significant difference was found between the amount of bycatch species between the two trap designs. Now that an overview has been given for the total composition of bycatch species it is time to get a closer look at the composition differences between the two trap designs. Table 7 gives an overview of the count and percentages of total most caught bycatch species divided over trap design.
What can be seen clearly is that the count of most species is higher in the traditional traps than in the funnel traps, what indeed confirms what the statistics have shown. However, there are exceptions in the count of the vermillion snapper where both traps (119 in traditional traps to 112 in funnel traps) are almost the same. Also, the funnel trap caught more blue tangs than the traditional trap did (25 and 18 respectively). In traditional traps 77,69% of bycatch comprised of redfish species while 78,48%
of the funnel traps bycatch comprised of redfish species. Graph 7 compares the frequency counts of the bycatch species per trap. Considering frequencies, three of the four redfish species (vermillion snapper excluded) show a clear decline when using funnel trap instead of the traditional trap.
Table 7. Overview of frequency of the eleven most caught species as bycatch in the lionfish traps divided over trap design
Species Count traditional Perc. (%) Count funnel Perc. (%)
Banded Butterflyfish 26 1,92 20 2,87
Blackfin Snapper 685 47,01 326 46,77
Blue Tang 18 1,24 25 3,59
Coney Grouper 29 1,99 8 1,15
Red Hind 136 9,33 55 7,89
Spotted Goatfish 37 2,54 12 1,72
Longspine Squirrelfish 107 7,34 56 8,03
Vermillion Snapper 119 8,17 112 16,07
White Grunt 72 4,94 21 3,01
White-spotted Filefish 31 2,13 8 1,15
Yellow-eye Snapper 195 13,38 54 7,75
Graph 6. Pie chart showing the total bycatch species composition of the eleven most frequently caught species caught in both trap designs. The four largest slices are all the common redfish species of snappers and the red hind and comprise 78,09 % of the total bycatch.
29 Results from scoring system
A significant difference was found between the effectiveness scores and trap design (P=0.000). As can be seen in graph 10, the traditional trap had and mean effectiveness score of 4,73 while the funnel trap had a mean score of 5,17.
Graph 7. Bar-chart comparing the bycatch frequency count between the two trap designs (funnel in orange and traditional in red). The four redfish species (snappers and red hind) are put on the left side of the x-axis. An * behind the species name show the possible significant differences.
Graph 10. The mean effectiveness score per trap design (4,73 for traditional and 5,17 for funnel trap).
* *
*
*
*
30
Table 8 shows the distribution of scores for the two different trap designs. The high number of traps that scored a five can be explained by the number of traps that were hauled having zero lionfish and less than 10 bycatches. In total 212 traps did not have any lionfish resulting in a zero for lionfish score.
197 traps had less than 10 individuals of bycatch giving a bycatch score of 5.
The traditional traps had a total of 48 traps scoring below a score of five. In contrast, the funnel trap only had 14 traps scoring below a score of 5. When looking at the highest scores (having a score higher than 5), the traditional trap had a total of 25 traps scoring higher than a score of 5. The funnel traps had a total of 27 traps scoring higher than a score of 5.
Table 8. Distribution of score results between the trap designs.
Effectiveness score Traditional trap score count Funnel trap score count
1 1 0
2 6 1
3 13 1
4 28 12
5 56 105
6 15 18
7 5 5
8 4 2
9 1 2
10 0 0
Interestingly from the scoring results is to see the differences in lionfish to bycatch ratio per haul. In total five traps caught at least one lionfish and had no bycatch at all. Two funnel traps caught two lionfish and had no bycatch. One funnel trap caught one lionfish and no bycatch. One traditional trap caught two lionfish in total having no bycatch. And one more traditional trap caught one lionfish and no bycatch. Of the 28 traps that had 21 or more bycatches, only 2 were a funnel trap.
31