The following article summarises the contents of a thorough assessment of the impact of European Union purse seine fisheries on marine turtles in the Indian Ocean (and Atlantic Ocean), prepared by Clermont et al. and submitted to the September 2012 meeting of the Indian Ocean Tuna Commission’s ‘Working Party on Ecosystems and Bycatch’ (IOTC-2012-WPEB-35).
The WPEB document bears the title:
EU PURSE SEINE FISHERY INTERACTION WITH MARINE TURTLES IN THE ATLANTIC AND INDIAN OCEANS: A 15 YEAR ANALYSIS
Sandra Clermont, Pierre Chavance, Alicia Delgado, Hilario Murua, Jon Ruiz, Stéphane Ciccione and Jérôme Bourjea
With permission of the authors, I have condensed the main findings and conclusions of this 74-page report to form an “extended abstract” in order to focus specifically on issues pertaining to the Indian Ocean. The full paper (2.2 MB) – including details of the methodology, numerous tables and figures, additional information relating to purse seine fisheries in the Atlantic Ocean, and references – can be downloaded from the IOTC webpage (www.iotc.org) and Bibliography Resource of the IOSEA website.
The report’s main conclusions with regard to the Indian Ocean can be further distilled, as follows:
- Based on available observer data, the average annual bycatch of marine turtles in fishing sets of European Union (EU) purse seines operating in the Indian Ocean is estimated to be in the order of 250 animals. About 3/4 of these by-caught turtles are released alive, suggesting that the number of marine turtles killed in the Indian Ocean EU purse seine fishery is in the order of 60 individuals per year. This finding is consistent with studies in other ocean basins showing very low rates of turtle mortality in purse seine operations.
- The observed impact of the EU tropical purse seine fishery is extremely low in comparison to other worldwide estimates of turtle mortality in industrial and artisanal fishing gears – such as pelagic long-lines, gillnets, and trawl nets – which are associated with estimated mortality rates that are several orders of magnitude higher.
- However, it is worthwhile noting that observations on sets do not take into account the ghost fishing phenomenon occurring on floating devices (some of them being lost by owners) that do not end up in a fishing set. Thus, the total impact of Fish Aggregating Devices (FADs) in the Indian Ocean was not assessed and is unknown. FADs that are deployed, not observed and not recovered (in whole or in part) are believed to be an unquantified source of turtle mortality through entanglement and subsequent drowning.
- Modifications in FAD design that aim to reduce the entanglement of turtles could begin to mitigate the problem of “ghost fishing” by unmonitored drifting FADs.
A higher level of observer coverage in the purse seine fishery would probably improve by-catch estimates, and availability of data on the number of FADs actually deployed would provide a better indication of the total impact of this gear type. However, the present comparison between observed purse seine mortality rates with those of other industrial and artisanal fishing gear – such as long-lines, gillnets and trawl nets – points to the urgent need to prioritise the latter for more intensive investigation and mitigation action.
Extended Abstract of the Clermont et al. WPEB Paper
Five species of marine turtle are present in the Indian Ocean: the green turtle (Chelonia mydas), the hawksbill turtle (Eretmochelys imbricata), the loggerhead turtle (Caretta caretta), the olive ridley (Lepidochelys olivacea) and the leatherback turtle (Dermochelys coriacea). Bycatch of marine turtles, vulnerable or endangered species, is a growing issue of all fisheries, including the oceanic purse seine fishery which is responsible for 40% of the total tuna catches in the Atlantic and Indian Oceans (ICCAT, 2011; IOTC, 2011). This fishery represents around 500,000 tons of tuna per year – primarly yellowfin, skipjack, and bigeye – and is largely dominated in the Atlantic and Indian Oceans by European Union (EU) fleets composed of Spanish and French vessels.
Purse seining for tuna involves surrounding the tuna school with non-meshing net about 1,500 m long. It can take place on a free swimming school (FSC) or on a school aggregated under a floating object, called a Fish Aggregating Device (FAD). The FAD can be natural objects, such as logs or palm branches, or man-made objects with a buoy incorporated to indicate their position. The EU has developed fishing under FADs since 1993.
Many reports and grey literature have already emphasized that the purse seine fishery has a low bycatch level. Hall (2012) noted that marine turtle bycatch is usually less than 1% of the sets, with captures numbering generally one individual and, in the vast majority of the sets, the turtle is released alive. Most of the bycatch occurs when the purse seiners encircle the tuna schools. Marine turtles are most of the time encircled in the net and are freed alive when the net is pulled up from the water towards the power block. Turtles are also attracted by floating devices such as FADs which usually have pieces of nets hanging below them. A turtle may become entangled for a long time and mortality may occur by drowning. An unknown percentage of drifting FADs get lost due to currents, creating the phenomenon of “ghost fishing by drifting FAD” (Chanrachkij & Loog-on 2003).
The present paper seeks to assess marine turtle bycatch at a spatial and temporal level in the European purse seine fishery operating in the Atlantic and Indian Oceans. As bycatch is poorly reported in fishery log-books, research is usually carried out using observer program data (Rochet & Trenkel, 2005). In the purse seine fishery in the Atlantic and Indian Oceans, marine turtle bycatch is reported only by onboard observers, and that observer coverage is low.
Such a low coverage contributes to the difficulty in producing solid estimates of marine turtle bycatch and mortality due to purse seine activity (Sanchez et al., 2007). Amandè et al. (2012) stated that the low observer coverage in the EU purse seine fishery in the Indian Ocean resulted in large uncertainties in bycatch estimates: up to 50% of mean square error. Nevertheless, since 2007 the programs in both oceans have reached at least the 5% coverage recommended by ICCAT and IOTC.
In the present study, we carried out a tentative elevation of the observed bycatch data to the total fishing effort per year and ocean in order to have an order of magnitude of the total number of marine turtles accidentally caught by the purse seine fishery in the Atlantic and Indian Oceans. We used information from French and Spanish fishing statistics from logbooks to determine a raising factor based on the effort of the fleets (number of sets on FSC and FAD per 1° square, per year and per quarter). In this way, we established an estimation of the total marine turtle bycatch based on information during observed fishing sets on FSC or FAD. Since there is no available information on the total number of FADs deployed by the EU purse seine fishery, it was not possible to estimate the real impact of ghost fishing by drifting FADs.
The study was based on data collected by French and Spanish observer programs from 1995 to 2011, a period where more than 115,000 fishing sets were realized by the EU fleets in the Indian Ocean. A total of 6,145 fishing sets were observed in the Indian Ocean, including 3,132 on drifting Fish Aggregating Devices (FAD) and 3,013 on Free Swimming Schools (FSC), representing observer coverage of 5.1%. The fishing effort is concentrated in the Mozambique Channel during the end of the first and all of the second quarter of the year, before moving to the northwestern Indian Ocean in the third, fourth and beginning of the first quarter.
At the temporal and spatial level, data were organised and analysed by fishing mode (FAD vs. FSC) as well as by year, quarter and statistical square of 1°. In order to evaluate the impact of the fishery, bycatch distribution was compared to the total fishing effort of the EU fleet, as well as to the known marine turtle post-nesting migration routes, nesting population abundances and known feeding areas. The species composition, the size and sex structure of bycatch were also assessed.
The following figures show the spatial distribution of the total fishing effort (in number of sets per 1° square) from 1995 to 2011, as well as the set observation effort. The spatial coverage of the observer programs in term of sets on FSC (shown in the full paper) seems to contribute to a good coverage of the whole fishing area and effort; whereas the observer coverage for FADs (also shown in the full paper) is over-represented in the Mozambique Channel compared to the North-Western Indian Ocean. There are currently no available data on the number of deployed FADs per fleet. Nevertheless, by comparing the fishing effort on FADs to the observation of object coverage (i.e. Figure 2a), the object observation effort seems to cover well the total fishing effort on FADs, even if the Mozambique Channel remains over-observed.
Figure 2a. Total fishing effort (on FADs) per statistical square of 1° of French and Spanish fleets from 1995-2011 in the Indian and Atlantic Oceans
Figure 4. Total object observation effort (in number of observations) per statistical square of 1° of French and Spanish observer programs from 2003-2011 in the Indian and Atlantic Oceans.
Over the study period, 182 turtles were caught during set observations in the Indian Ocean, 77% being released alive at sea. This translates to 20 dead turtles and a similar number of unknown fate. At the same time, from 2003 to 2011, 9,349 specific observations were carried out on floating objects whether they ended in a set or not. 238 marine turtles were observed upon which 73% were already free or entangled alive and therefore released alive. Thus, the percentage of marine turtles released alive is very similar between FADs and FSCs.
In the Indian Ocean, more than 81% of the by-caught turtles were identified by species and 88% were associated to a fate. With 58 observations, Lepidochelys olivacea was the most observed species, followed by Eretmochelys imbricata and Chelonia mydas (37 and 32, respectively). Only two leatherback turtles were caught during sets observed in the Indian Ocean from 1995 to 2011.
The sex, size and the life stage were determined for 117 of these by-caught turtles of which about 75% were juveniles. We hypothesize that an abundance of juvenile marine turtles in their drifting pelagic phase may be more attracted to FADs looking for protection or food, rather than just drifting. The species composition for marine turtles observed around floating objects was similar to that observed during fishing sets.
The data seem to indicate that the level of bycatch from FADs is higher in the Northern Indian Ocean and northwest of the Mozambique Channel. The highest capture rates on FADs and FSCs happen around India, but are low in the Mozambique Channel despite higher observation effort. In general, captures per set rarely amount to more than a single individual most of the time.
In the Indian Ocean, the olive ridley is clearly found more often in the northern area (N=58). It is highly likely that these by-caught individuals are from the highly threatened Indian stock, although a genetic analysis is needed to confirm such hypothesis. Hawksbill (N=37) and green (N=32) turtles are found throughout the coverage area. The latter two species are also the only ones by-caught in the Mozambique Channel.
A similar pattern emerges for interactions of marine turtles with floating objects. The highest turtle observation rates are located in the northwestern zone and occur during the third and fourth quarters. Hawksbill (N=40) and green (N=37) turtles were observed more often in the southern area, even though their interaction with EU purse seine fishery occurs throughout the fishing area. Olive ridley (N=74) observations on objects are clearly located more in the Northern Indian Ocean.
Figure 9: Number of observed marine turtles by species caught by French and Spanish purse seine fleets during observed sets on FADs and FSCs in the Indian Ocean for the period 1995-2011: (CCC = Caretta caretta, CMM = Chelonia mydas, DCC = Dermochelys coriacea, EIM = Eretmochelys imbricata, LOL= Lepidochelys olivacea.)
Based on observation of marine turtle by-catches on sets, we estimate that about 2,000 marine turtles were accidentally captured by the EU purse seine fleet in the Indian Ocean from 2003 to 2010, with a corresponding annual bycatch rate of 250 individuals (SD=157), with 77% being released alive. Even with large standard deviation due to the low observation rate, we can roughly consider that this fishery kills less than 60 individuals per year in the Indian Ocean.
However, it is worthwhile noting that observations on sets do not take into account the ghost fishing phenomenon occurring on floating devices (some of them being lost by owners) that do not end up in a fishing set. Pieces of net hung below the FAD are believed to be the cause of marine turtle mortality by entanglement and subsequent drowning. The presence of turtles under a FAD is difficult to observe from the surface by on-board observers, whereas the chance of observing a turtle increases when a set is done under a FAD, since the marine turtle will be by-caught in the net.
It is important to remind that because of the low level of bycatch occurrences, the large spatial distribution of the fishing activity, and the complex behaviour of all marine turtle life stages, the present dataset does not allow an accurate and reliable total bycatch estimates. In order to improve the sample collection, it is highly recommended that higher observation coverage should be achieved and species identification improved.
Some developments and/or modifications of the FAD design can be made to mitigate the entanglement of turtles. If we consider the potentially high number of FADs deployed and the cryptic mortality they are believed to cause (not evaluated here), it is recommended to develop FADs without pieces of nets hanging below the FAD, or alternative models without nets, since the mesh size of these net fragments used in such FADs appears to be a key contributing factor for marine turtle entanglement (Amandè et al., 2008). The IOTC Working Party on Ecosystems and Bycatch has already recommended that FADs be designed without nets hanging underneath and using biodegradable materials (IOTC Resolution 12/04).
Finally, it is useful to compare the impact of the EU purse seine fishery on marine turtles in a wider context and to highlight that other fisheries, such as artisanal fisheries, may have a greater impact on marine turtles than industrial fisheries. A recent study showed for example that the annual turtle catch in the south-western province of Tulear (Madagascar) alone is between 10,000 and 16,000 (Humber et al., 2010). It is also important to note that currently it is estimated that 30 % of the tropical tuna catches in the Indian Ocean are done by gillnets, which are also thought to have a high level of bycatch.
Therefore, it is necessary to investigate and implement pilot observer programs in those fisheries, for which a complete lack of data exist, in order to globally evaluate different fishery interactions with turtle populations and associated turtle mortality. Lastly, we must keep in mind the impacts of fisheries in the light of other land-based or coastal threats. It is clear that, despite strong legislation prohibiting the direct take of turtles throughout the region, this is still regarded as the most important threat (see review for the Indian Ocean in Bourjea, 2012).