Fads Fisheries Development in the WECAFC Region and the Impact on Stock Assessments

FADs fisheries development in the WECAFC region and the impact on stock assessments

Nelson Ehrhardt, Julie E. Brown and Bruce G. Pohlot Rosenstiel School of Marine and Atmospheric Science University of Miami

Merida 3 November 2017 FADs are the most significant gear development in the world The use of FADs is an irreversible process … are here to stay

Figure 1. Geographical distribution of drifting FADs in tuna purse seining fisheries (Left created using IATTC public domain data, Right from Maufroy et al. 2015)

Most FAD developments are Ad Hoc We know fish gather around FADs but we do not know with certainty why they exhibit such behavior

Figure 2. Objects (natural or man-made) attract fish in the pelagic environment (pictures from several sources), yet the cause(s) are still poorly understood relative to FAD enhanced fishing. In the pelagic environment, sea surface is the only physical reference and light dominate Predator prey condition, and vision has a unique and significant role for prey and predators

Figure 3. Light incident angles and projected fish vision (left), light wave depth penetration (center) and vision of light refraction from a point under FADs. Example: Sailfish activity governed by light

Figure 4. Photokinetic response of Pacific sailfish activity to day light and full moon light (From Pohlot and Ehrhardt 2016). Example: Tuna daily activities governed by light

Figure 5. Daily pattern of aggregation, figure from Govinden et al. 2010. Example: daylight activities governed by avoiding light in association with FADs

Figure 6. Day and night distribution of fish resources aggregated under moored FAD in the eastern Caribbean and as detected by hydroacoustics. Left night distribution, Right day distribution (from Doray et al. 2005). 1. Food chain hypothesis proposes that large, predatory fish are actively predating upon the smaller fish, which are aggregating at the floating objects, and thus an entire food chain builds up with the FADs as it’s foundation.

The hypothesis was originally proposed for the mahi, which also associate with FADs, but was extended to tuna and marlin.

The large diversity of fish that are found associated with FADs would seem to support this theory (Amandè et al. 2010), however there is little evidence that there are enough small bait fish present to support a 20-40 ton school of tuna, which need to eat approximately 5% of their body weight per day (Olson and Boggs 1986).

Stomach contents of tuna associated with floating objects show that they are predominantly feeding on species that are not associated with the floating object (Malone et al. 2011). 2. The indicator-log hypothesis: fish of all sizes/trophic levels are attracted to natural floating objects because they tend to accumulate in waters that are relatively rich.

Frontal zones in the vast pelagic environment represent areas suitable for the “triad” (enrichment, concentration, retention) of physical ocean properties that are advantageous to fish (Bakun 1996).

One indicator that this hypothesis applies to tunas is the observation that tunas associated with FADs artificially seeded in areas that are considered low-productivity, are in poorer condition (empty stomachs, lower fat content) than their free-school counterparts (Hallier and Gaertner 2008, Robert, Dagorn et al. 2014).

This suggests that the fish are being “tricked” into thinking that they are being drawn or led into rich waters, when they actually are not, also known as the ecological trap theory.

The higher proportion of empty stomachs in FAD-associated tunas (Ménard, Stéquert et al. 2000) can be interpreted as evidence of the ecological trap (Table 1) . Fish collected in FADs show significantly less food in their stomachs

Impact on feeding efficiency due to changes in feeding dynamics has to have an effect on population fecundity and predation regimes

Table 1. Feeding Efficiency of associated and unassociated tunas, from Menard, Stequert et al 2000. 3. The meet up point hypothesis, proposed by Dagorn (1999). Fish seek out anomalies in the pelagic environment in hopes that others will do the same, and subsequently form larger schools. This has been described as an enhanced encounter rate.

Commercial data and Fontenenau (2013) confirm that FAD associated schools of tuna are larger than free-schools.

FAD associated schools have a larger range of sizes compared to the relatively homogenous sizes of fish in free-swimming schools (Wang et al. 2012) suggesting that the mechanism for schooling around a FAD is either different, or the impulse is stronger than fish that are free-swimming, either as individuals or as “sub optimal” size schools.

FAD caught tuna are smaller than free schooling tuna, and this phenomenon is particularly conspicuous for skipjack and yellowfin.

Acoustic telemetry experiments performed on bigeye scad endorse the hypothesis that fish are more likely to arrive as individuals or small groups, and leave the FAD in larger groups (Soria et al. 2009). 4. The shelter from predators hypothesis applies to intranatant species that prefer to stay very close to the FAD.

In the three dimensional environment, it is impossible to be looking in all directions at once. By staying very close to an object, there is one less direction to monitor for predator attack. This theory applies to all structures, not just floating ones.

5. The seeking shade hypothesis also applies to intranatant species that remain very close to the FAD.

Helfman (1981) proved that observers in shade can see an illuminated object 2.5 further than an illuminated observer could see an object in shade. Clearly, the vision is improved in the shade. Further evidence that some fish may take advantage of the change in visibility is the fact that dark colored fish have higher preference for being directly under the FAD than light, silvery fish (Hunter 1966). The darker fish may gain a greater camouflage advantage in the shade. Clearly, the affinity for certain fish to associate with floating objects, and subsequent changes in daily or seasonal behavior affects the overall ecology of the pelagic environment.

Vis-à-vis, the surrounding ecology and presence of others appears to change the behavior and probable motivations for individual fish associated with the floating object.

For target species like tuna, behavioral changes are likely to be significant, and there is a need to factor this into estimates of catchability for fisheries assessment.

In sum, FADs do alter the general ecology of the pelagic environment that favors the Increase in biomass density and by consequence, an increase in catchability.

Density enhanced fishing in the WECAFC: a) the condominium or “Cuban casitas” used to gather spiny lobsters in such artificial habitats for divers to retrieve them, b) spiny lobster trap fisheries that use conspecific attraction behavior by using sublegal lobsters as bait in commercial traps to attract legal-sized adults, and c) the use fish attracting devices (FADs) in costal migratory pelagic fisheries. New advanced satellite hydro-acoustic technologies transforming “fishing into harvesting”

Figure 8. Example of communication web. Adapted from Satlink documentation. (www.satlink.es) Sun energy driven satellite hydro-acoustic transducer used with FADs to assess biomass Technologies allow fishers to “see and check” status of biomass aggregation under FADs

Figure 9: Echogram output from Zunibal's Tuna8 Explorer buoy and software. Courtesy of Zunibal.com Keeping track of the dynamics of biomass accumulation under FADs and associated oceanography

Figure 10: Example of visualization software from Marine Instruments providing environmental and oceanographic information along with maps and echogram outputs. Courtesy of MarineInstruments.es Mathematical modeling uisng FADs hydro-acoustic data used to forecast catch

Figure 11. Model expressing relationship between actual catch (y-axis) at a FAD versus the amount of biomass using modelled data (x-axis) published by Lopez et al. 2016) from an echo-sounder equipped FAD. What is the issue with FAD statistical data and stock assessments?

Catch per unit of effort is an index that is directly proportional to relative abundance

퐶푡 = 푞푁ഥ푡 푓푡

Stock assessment methods attempt to estimate the proportionality factor “q” from fishery data

However, CPUE from FADs is not directly proportional to relative abundance, it is an index of local (not population) density concentration under FADs

Such condition is termed hyper stability of CPUE

Consequences: over estimate abundance and under estimated fishing mortality (OVER OPTIMISTIC VIEW OF THE STATUS OF THE RESORUCES) CPUE = q * N

q = Catchability Page 380 N =66 Averageth Gulf and CStockaribbea Abundancen Fisheries Institute

Through a combination of acoustic telemetry, conventional tagging, and direct observations of spawning behavior, we have discovered that: i) All Nassau grouper aggregate every year (Semmens et al. 2005), ii) Older (larger) fish aggregate for a longer period of time -- a 50 cm fish may spend 130 hours on the aggregation, while a 70 cm fish would spend 230 hours (Semmens et al. 2007, Semmens et al. unpublished data), iii) Fish travel long distances during an aggregation, even if the FSA is not far from their home territory, with some fish swimming up to 230 km around an island < 20 km in length (Semmens et al., unpublished data), and iv) Smaller aggregations aggregate for a longer period of time, with a 500-fish aggregation potentially occurring for as much as 125 hours (5 days) longer than a 3500 fish aggregation (Semmens et al., unpublished data)

Hyperstability has been defined as CPUE staying the same as abundance declines, essentially a “masking” effect on population trends, and has been attributed to fish Figure 1. The relationship between true abundance and continuing to pack themselves at the same density but the catch-per-unit effort. Under standard model assumptions, mechanisms for hyperstability to this point have been CPUE is assumed to be proportional to abundance. Hyper- inferred. Here we provide behavioral evidence for how stability occurs when catchability increases with decreasing hyperstability manifests for aggregating species. Our data stock size. Hyper-depletion, where q drops faster than the population and therefore results in an under-estimate of indicate that for aggregating species like Nassau grouper, abundance, is shown for contrast. hyperstability is not simply a density and distribution issue as has been previously described for other species, but Semmens, B.X., S.A. Heppell, P. Bush, B. Johnson, C. McCoy, C. rather a life history issue. Behaviorally-mediated vulnera- Pattengill-Semmens, and L. Whaylen. 2007. An intra- and inter- bility issues provide a potential mechanism for depensation annual analysis of Nassau grouper size distributions from a recently (Allee effects) under the guise of hyperstability. Finally, protected spawning aggregation in the Cayman Islands. Proceedings of the Gulf and Caribbean Fisheries Institute 58:585-586. based on these results we conclude that: Walters, C. and J-J. Maguire. 1996. Lessons for stock assessment from the i) Protecting only some (but not all) regional FSAs northern cod collapse. Reviews in Fish Biology and Fisheries 6:125- will not protect fish from overexploitation, and 137. ii) Protecting only the FSAs themselves and not the . corridors to and from the FSAs will not protect fish from overexploitation.

LITERATURE CITED Erisman, B.E., L.G. Allen, J.T. Claisse, D.J. Pondella II, E.F. Miller, and J.H. Murray. 2011. The illusion of plenty: hyperstability masks collapses in two recreational fisheries that target fish spawning aggregations. Canadian Journal of Fisheries and Aquatic Sciences 68:1705-1716. Hilborn, R. and C.J. Walters. 1992. Quantitative Fisheries Stock Assessment: Choice, Dynamics and Uncertainty. Chapman and Hall, New York, New York USA. 570 pp. Rose, G.A. and D.W. Kulka, 1999. Hyperaggregation of fish and fisheries: how catch-per-unit-effort increased as the northern cod (Gadus morhua) declined. Canadian Journal of Fisheries and Aquatic Sciences 56:118-127. Semmens, B.X., K.E. Luke, P.G. Bush, C.V. Pattengill-Semmens, B. Johnson, C. McCoy, and S. Heppell. 2005. Investigating the reproductive migration and spatial ecology of Nassau grouper (Epinephelus striatus) on Little Cayman Island using acoustic tags – An Overview. Proceedings of the Gulf and Caribbean Fisheries Institute 58:199-206.

Directly proportional relationship

Axis with an index of relative abundance not affected by density enhancing devices

퐶푃푈퐸 푞 If b=1, then the ratio of the CPUEs is a measure of 퐹푆 퐹푆 ഥ 1−푏 = 푁 Fishing power differences “independent of abundance” 퐶푃푈퐸퐹퐷 푞퐹퐷 Another interpretation of CPUE hyper stability: the rate of null sets in purse seine tuna fisheries decreases much slower in FAD fisheries than in free school fisheries due to the FAD aggregating power 1

0.9

a i

c 0.8

D 0.7

o 0.6

l 0.5

f s

s y = 0.3932x + 0.6979 e

c 0.4 R² = 0.3419

o 0.3

i 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

t a R Ratio of successful sets on SCHOOLING FREE

Figure13. Fraction of successful purse seine sets on FAD associated on fraction of purse seine sets on free schooling tuna in the Indian Ocean (Adapted from data in Floch et al. 2012). OPPORTUNITIES FOR RESEARCH OF THE THE CPUE HYPER STABILITY AND IMPROVE FISHERY STATISTCS

Table 2. Fish aggregations at moored FAD in Martinique, sampled using echo-sounder buoys, underwater video, trawl fishing, and longline fishing. Adapted from Doray (2007).

Species Location Abundance 1 Small tunas (blackfin, yellowfin, skipjack, frigate) Surface layer (0-10 m ) Little biomass (2%) average 20 cm

2 Wreckfishes (jacks, triggerfish, dolphinfishes) 20 m close to mooring line Little biomass (2%)

3 Larger tunas. Mostly blackfin 50 cm, skipjack and 30-100 m 95% of estimated biomass yellowfin also. from echosounder

4 Scattered large predators, blue marlin mostly. subsurface Low biomass (1%)

5 Large plankton Above thermocline, widely Density varies with season spread UNDERSTANDING BEHAVIORAL ASPECTS OF FISH ATTRACTED AND AGGREGATED UNDER FADS

Figure 20. Diel behavior of major fish assemblages under mFADs. Figure from Doray et al. (2007). Multi-gear multi-species fisheries integrated in statistical system Leading to differentials in fishing efficiencies affecting CPUE

Figure 21. Per trip, FAD fishing yields higher landings and daily wages, however value added / crew member / hour is higher for inshore fishing activities because the trips are considerably shorter. (Guyader et al. 2013) CPUE for given species are effected by spatial FAD deployment effects

Figure 22. Locations of known (yellow dots) and approximate (green dots) mFADs in South Haiti in 2015, and landings ports (red). Valles (2015). Table 5. Landings are highly variable, by location and season. For some ports, there are considerable trips returning with zero fish. Valles (2015). Figure 24. Imports of cheap, foreign seafood directly competes with local fishery products in the market. Mathieu et al. 2014. Impact of interacting fishing intensity (gear density) on CPUE triggered by demand

Figure 23 The FAD fleet began to reach capacity around 2001 for Martinique and Guadeloupe, Mathieu et al. 2013.

Page 514 66th Gulf and Caribbean Fisheries Institute

Figure 5. Species captured around FAD in the three islands, Guadeloupe, Dominica and Martinique in 2010. Source: SIH and Fisheries Division of Dominica.

Page 514 66th Gulf and Caribbean Fisheries Institute

Page 512 66th Gulf and Caribbean Fisheries Institute

Although the three islands have similar type of fleet, the 16 in Martinique whereas more than 400 FADs are present FAD development is peculiar to each island and fishing in Guadeloupe (data from SIH, 2013). An airplane survey behavior differs. The FAD development on the French conducted in 2008 above Martinique and Guadeloupe islands went through tree stages (Figure 1): could already show this singularity (Figure 3). Fishermen i) A light rise where Martinican FAD activity is in Dominica and Martinique visit only one or two FADs above Guadelupian for the first seven years of per fishing trip whereas in Guadeloupe they will visit on FAD fishery (1985 - 1992), average 4 to 5 FADs in the same trip thanks to their close ii) A strong increase where Guadeloupian FAD proximity. activity goes over Martinican between 1992 to The effect and impact of FADs utilization mode on the 2002, and catches can explain a part of this difference between FAD Figure 6. Fiiis) hAe cremilinegn a roaunvde 3r0a0 gveess eals gfoer bowthh iselannd s tahs eify sftiashritn g FinA MDar tinique and that same activity in Guadeloupe. a maximum capacity of the FAD fleet has been Indeed, fishermen from Guadeloupe mainly target dolphin activity betweenre a1ch9e8d 5in a 2n0d02 2i0n 0G5u aidne loFurpee nacndh 2I0s0l6a nind s. fish. For that reason, thFerieg aurer ae l ot8 o.f pLrivaanted FiAnDgss o wpneedr trip and per gear in Guadeloupe. Source: SIH. Martinique. by one fisherman wSheoreuasr cien : MOa.rt iGniquuye athdee rfi sehitn ga l. 2011. Figure 5. Species captured around FAD in the threea cistilavnitdys ,i sG muaodreel oourpgea,n Dizoemdi nbicya g arnodu p s. The consequence is If we compMaraer titnhieq uneu imn b2e0r1 0o.f S Foruerncech: SFIHA Dan dv eFsisehlesr ietos Divistihoant offe Dwoemr iFniAcaD. s are deployed in Martinique compared to Dominican FAD vessels for the past 5 years, we can see Guadeloupe. Martinican fishermen target mainly blue that the situation is completely different. Dominican FAD marlin and yellowfin tuna. The composition of landings is Mathieu, H. et al. GCFI:66 (2014) Page 513 fleet is still in full growth. From 2009 to 2012 we had an different from one island to the other (Figure 4). between 15% and 25% (data from Reynal et al. 2013). In mately of FAD vessels are exclusive in the FAD fishing. additional 100 FAD vessels (2F00i8g aundr e20 120 )i.n Guadeloupe, the three main species, The fish poItns reMpresaenrt taibnouitq tuhee s,a mien p ro2po0rti0on9 i n and 2010, blue marlin, main Nowadays a clear differeyenllocwefin btuneat, wdolepheinn fi sth haned btlhuer meaerli ni wselrae n36d%s to Guasdpeloeupcei eansd feviesnh meodre, 3r4e%p, rine sMearntintiequde. 4In 4% to 36% of the total FAD 35%, 33% to 29% and 18% to 9% respectively of total Guadeloupe the most common strategy applied by FAD is the FAD density presentF ADisn l andtihnges (dastae faro.m GMuyaaderrt eitn ali. q20u10e). Fiananllyd in fishcerametnc ihs teo sus,e mruletipsuprpeosce tfiixvede gleayrs. in aAdditti on ttoh e same time, dolphin fish Dominica we obtained in 2010 32% for the yellowfin tuna, the FAD activity (represent 31% of the FAD vessels), see Dominica have about the sa16m% efo r tnheu domlphbine firsh aondf 7%F fAor tDhe bliune m atrlhine (diarta Figucrea 7t. ches were between 9% and 4% and yellowfin tuna from Fisheries Division database, 2013). See Figure 5. Faced with these results, the question is: what is the waters, 29 in Dominica (DefoeB aasnicadlly ,P whe fiolubnde thret s,a m2e0 sp1ec3ies) ,w earer coaputurnedd in attractiveness of FADs compared to the other activities? the three islands but in different proportions. Main species Thanks to studies carried out in Guadeloupe, we were captured are the yellowfin tuna (represented in yellow in able to note several observations. In terms of quantity Figure 5) with the highest number of catch in Guadeloupe, landed, FAD fishing and trolling line seem to be very with decreasing numbers toward the south, same for the attractive compared to other fishing activities such as pots dolphin fish (represented in light green in Figure 5), or gillnets (coastal activities) even if there is an important another important species captured on FAD and Finally the variability factor in FAD fishing (Figure 8). In average, blue Marlin (represented in blue in Figure 5) is more FAD fishing induces bigger gross revenue than coastal Figure 6. Fishermen averacgapetur eda ing tehe Sowuth (eMnar tintihquee)y fo r 2s0t1a0.r t FAD activities (figure 9). However when you deduce the costs of FAD fishing started with young fishermen in the operation (fuel, ice, food, gear, maintenance…) the income activity between 1985 and 2005 in French Islands. foFr tihge FuArDe fi sh8er.m aLn ias nclodseirn bguts st ilpl aeborv e ttrhiep in caomne d per gear in Guadeloupe. Example of integrated multiFren-chgear island soperations, their first results in terms of quantities of Source: SIH. fish landed most likely convinced older fishermen to coSmiongu frocme f:is hO po.t sG anud ygiallndetes (rF iegutr ea 10l.). 2Al0so1, i1t i.s practice that type of fishery (Figure 6). Indeed, FAD important to point out that the variability for FAD fishing fishing looked as a productive fishery compared to coastal can put the fisherman in a dangerous economic situation in fisheries such as fishing pots practiced at that time by a a short period due to the cost of the associated with a majority of fishermen. This change of fishing behavior fishing trip. Finally, if you take in consideration the size of from older people could possibly have an impact on coastal the crew and the hours fished, fish pots and gillnets may resources by exploiting them less. appear in some situations as more interesting business than FAD and trolling line on a monetary approach (Figure 11). Fishing Activities in Addition to FAD Fishery, Where From this, one can easily understand why the choice of the Effort is Allocated FAD activity and the intensity of this activity are not The increase in FAD fishing does not necessarily homogeneously distributed within the fleet. Non-monetary means a desertion of the other fishing activities. For incentives can also be seen to influence fishermen’s Figure 1. Number of FAD evxaemsplse ein ltshe Ffrreoncmh is la1nd9s i8n 520 10t,o on ly2 200%0 9ap prionxi - behavior (Guyader et al. 2013). Martinique and Guadeloupe. Source: SIH Ifremer (“Système d’Information Halieutique”).

n = 3000 trips, 105 economic surveys (year 2008)

Figure 7. FPigruorep 7o. rPtriooponrt ioon fo fv veesssselse plser fpisehinrg faicstihvitiyn agm oangc tiviFtiyg uarem 9.o Gnrogss revenuFe ipgeru trripe a n9d .p eGr greoars ins Grueadvee- nue per trip and per gear in Guade- FAD fleet in Guadeloupe and Martinique 2010. loupe. Source: O. Guyader et al. 2011. FAD fleet iSno uGrceu: aSIdH.e loupe and Martinique 2010. loupe. Source: O. Guyader et al. 2011. Source: SIH.

Figure 4. Landing structure of the catches around FADs in Martinique and Guadeloupe during dolphin fish season and out of the season, 2008. Source: SIH Obsdeb. Mathieu et al 2013

Figure 2. Number of FAD vessels from 2006 to 2011 in Figure 3. Airplane survey of FADs above Guadeloupe and Martinique and Guadeloupe; and from 2008 to 2012 in Martinique, the red spots representing the FADs. Source: Dominica. Sources: SIH and Fisheries Division of Dominica. Ifremer, 2008. Page 512 66th Gulf and Caribbean Fisheries Institute

Although the three islands have similar type of fleet, the 16 in Martinique whereas more than 400 FADs are present FAD development is peculiar to each island and fishing in Guadeloupe (data from SIH, 2013). An airplane survey behavior differs. The FAD development on the French conducted i n 2008 above Mart in iqMautheie u,a Hn. d et alG. u GaCdFIe:6l6o (u20p1e4) Page 513 islands went through tree stages (Figure 1): could already show this singularity (Figure 3). Fishermen between 15% and 25% (data from Reynal et al. 2013). In mately of FAD vessels are exclusive in the FAD fishing. i) A light rise where Martinican FAD activity is in Dominica2 00a8n adnd M201a0r tini nGiuqadueelo upvei, stihte tohrneel ym aoin nsepe coiesr, twToh e FfiAsh Dposts represent about the same proportion in above Guadelupian for the first seven years of per fishing ytrelilpow fwin htunear, edoalsph iinn fi shG aunda bdluee lmoaurlpin ew etreh 3e6y% two ill Gvuaidseilto upoe na nd even more, 34%, in Martinique. In 35%, 33% to 29% and 18% to 9% respectively of total Guadeloupe the most common strategy applied by FAD FAD fishery (1985 - 1992), average 4 toF A5D sF laAndDingss (idnata tfhroem Gsuaymadeer ett rail.p 2 0t1h0)a. nFiknasll yt ion thefisihre rcmleon sise to use multipurpose fixed gears in addition to Dominica we obtained in 2010 32% for the yellowfin tuna, the FAD activity (represent 31% of the FAD vessels), see ii) A strong increase where Guadeloupian FAD proximity. 16% for the dolphin fish and 7% for the blue marlin (data Figure 7. activity goes over Martinican between 1992 to The effefrcomt aFinshder ieism Dpiviasicont doatfa bFasAe, 2D01s3) .u Steiel Fiizguartei 5o. n mode oFanc etdh weit h these results, the question is: what is the Basically, we found the same species were captured in attractiveness of FADs compared to the other activities? 2002, and catches can tehex tphrelea isnla nads bpuat rint doifffe retnht ipsro pdoritfiofnes.r Menaicn esp ebciest ween TFhaAnksD to studies carried out in Guadeloupe, we were captured are the yellowfin tuna (represented in yellow in able to note several observations. In terms of quantity iii) A ceiling around 300 vessels for both islands as if fishing in MFaigrutrien 5i)q wuithe thaen hdig htehst anut msbaerm ofe c aatcch tiinv Giutayde lionup Ge, uadlanedleod,u FpAeD. fishing and trolling line seem to be very a maximum capacity of the FAD fleet has been Indeed, fishewritmh deecnre afsrinog mnu mGberus atodwearldo tuhep seo umth, asaimnel yfo rt athre geta tdtraoctlipveh cionm pared to other fishing activities such as pots dolphin fish (represented in light green in Figure 5), or gillnets (coastal activities) even if there is an important reached in 2002 in Guadeloupe and 2006 in fish. For thaatn ortehear simopnor,t atnht sepreceie sa craept uare dl oont FoAfD panrdi vFianatlely tFheA Dvsa roiabwilintye fdac tor in FAD fishing (Figure 8). In average, blue Marlin (represented in blue in Figure 5) is more FAD fishing induces bigger gross revenue than coastal Martinique. by one fischaeptrurmeda inn th e Swouhthe (rMeaartsin iquien) forM 201a0r. tinique the actfiviistihesi (nfigu re 9). However when you deduce the costs of activity is morFeA Do rfgishaingi zsteardte db wyi thg ryouungp sfi.s heTrmheen icn othnes eqoupeernatcioen (fiusel , ice, food, gear, maintenance…) the income French islands, their first results in terms of quantities of for the FAD fisherman is closer but still above the income coming from fish pots and gillnets (Figure 10). Also, it is If we compare the number of French FAD vessels to that fewer FfiAsh Dlasnd eadr em odst elpikleolyy ceodnv inicne d Moldaerr tif i nshieqrmueen tcoo mpar e dMa thiteuo, H . et al. GCFI:66 (2014) Page 513 practice that type of fishery (Figure 6). Indeed, FAD important to point out that the variability for FAD fishing Dominican FAD vessels for the past 5 years, we can see Guadeloupe. Martinican fishermen tabertwgeene 1t5% andm 25%a (idantac flaroynm pReuytnb atl hlete ua lf.e i2s0 h13e).r Imn an miante lay odf aFAnDg eversoseulss a ree ceoxcnluosivme inc thsei tFuAaDt ifioshnin gi.n fishing looked as a productive fishery compar2e0d08 taon dc 2o0a10s tian lG uadeloupe, the three main species, The fish pots represent about the same proportion in fisheries such as fishing pots practiced at thayetl lotwifmin etu nba, yd olaph in fish an d sbhluoe rmta rlpine wreiroe d36 %d uto e toGu atdheleou pce oasndt eovfen tmhoer e, a3s4s%o, cina tMeadrt inwiquiet.h I n a that the situation is completely different. Dominican FAD marlin and yellowfin tuna. The compos3i5t%i, o33%n t o o29%f anld a18n%f idtso hi9i%nn grge stprsecitp iv.ie lFsy io nf atoltlaly , ifG uyaodeulo utpae kthee imno sct oconmsmiodne srtraateigoyn a ptphlied sbiyz FeA Do f majority of fishermen. This change of fishiFnAgD s lbanedhinagsv (idoatra from Guyader et al. 2010). Finally in fishermen is to use multipurpose fixed gears in addition to fleet is still in full growth. From 2009 to 2012 we had an different frofmrom o onldeer piseolpalen cdou ltdo p otshsiebl yo htahvee arn i(mFpiaDgcotm uoinnirc aec wo e a4obstta)ain.le d in 201t0h 3e2 %c froer twhe yaelnlodw fitnh tuen a,h outrhse FfAiDs hacetidvi,ty f(riesphre sepnot 3t1s% aonf tdhe gFAilDl nvestssel s)m, seaey 16% for the dolphin fish anda 7p%p feora trh ei bnlu es omamrlien (sdaitau atiFoignusre 7a. s more interesting business than resources by exploiting them less. from Fisheries Division database, 2013). See Figure 5. Faced with these results, the question is: what is the additional 100 FAD vessels (Figure 2). In Martinique, in 2009 and 2010, bBlasuicaelly , wme fouand rthle isFanmAe, Ds pe camiens dwa etreri ocnalplt iunregd inl inea totrnac tiave nmesos onf eFAtaDrsy co ampaprerdo toa tcheh o t(hFer iagctuivriteie s?1 1). Thanks to studies carried out in Guadeloupe, we were the three islands but in differFenrt oprmopo rtihonis. ,M aoinn sep eccieas n easily understand why the choice of Nowadays a clear difference between the three islands species fisheFdish,i nrge Apctriveistiesn int eAddd i4tio4n% to FtAoD F3is6he%rcyap, t uWroed fhar ee rthehe yello wtfino tutna (rle preFsenAted iDn yel low in able to note several observations. In terms of quantity Figure 5) with the highest nuFmbAerD of caatccht inv Gituyad elaounpde, thlean deidn, tFeAnDs iftisyh ingo fa ndt htroislli nga clintie vsieteym toa rbee vneroy t the Effort is Allocated with decreasing numbers toward the south, same for the attractive compared to other fishing activities such as pots is the FAD density present in the sea. Martinique and catches, respeTchtei vineclreya.se iAn tF ADt hfeis hinsga dmoees nottd iolmnpheinc eefis,hs a(reidplryeso entledp ihnh oliigmnhto ggreeefn eiinos uFhisglu yre d5)i,s triboru gtilelndet s w(coiatshtailn a ctihvietie sf) leeveent .i f Ntheoren i-s man oimnpeorttaanrty another important species capitnurcede onnt FiAvDe san d cFiananll y tahel so vabrieab ilitsye feacnto r itno F AiDn filsuhinegn (cFeig uref i8s).h Ien ramveerange’, s means a desertion of the other fishing acbtliuve iMtiaerlsin. (rFeporerse nted in blue in Figure 5) is more FAD fishing induces bigger gross revenue than coastal Dominica have about the same number of FAD in their catches wereexa mbplee tinw theee Fnre nc9h %isla ndas nind 2 0140,% on ly a2c0nap%tdur ead pinp ythreeo Sxoluitl-ho (Mwartinbfiqeuiehn) afo vr 2i0ot1r0u . (Gnuay ader et aclti.v i2tie0s 1(f3ig)u.re 9). However when you deduce the costs of FAD fishing started with young fishermen in the operation (fuel, ice, food, gear, maintenance…) the income waters, 29 in Dominica (Defoe and Philbert, 2013), around French islands, their first results in terms of quantities of for the FAD fisherman is closer but still above the income Guadeloupe FAD fisheriesfish landed most likely convinced older fishermen to coming from fish pots and gillnets (Figure 10). Also, it is practice that type of fishery (Figure 6). IndeeDolphind, FAD Fishi mSeasonportant to point out that the variability for FAD fishing fishing looked as a productive fishery compared to coastal can put the fisherman in a dangerous economic situation in fisheries such as fishing pots practiced at that time by a a short period due to the cost of the associated with a majority of fishermen. This change of fishing behavior fishing trip. Finally, if you take in consideration the size of from older people could possibly have an impact on coastal the crew and the hours fished, fish pots and gillnets may resources by exploiting them less. appear in some situations as more interesting business than FAD and trolling line on a monetary approach (Figure 11). Fishing Activities in Addition to FAD Fishery, Where From this, one can easily understand why the choice of the Effort is Allocated FAD activity and the intensity of this activity are not The increase in FAD fishing does not necessarily homogeneously distributed within the fleet. Non-monetary means a desertion of the other fishing activities. For incentives can also be seen to influence fishermen’s example in the French islands in 2010, only 20% approxi- behavior (Guyader et al. 2013).

Dolphin Fish Off-Season

64 FADs deployed

Figure 1. Number of FAD vessels from 1985 to 2009 in Dolphin Fish season Dolphin Fish Off-season Figure 4. Landing structure of the catches ar ound FADs in Martinique and Guadeloupe during Dolphin Fish 50% 28%Figure 4. Landing structure of the catches around FADs in Martinique and Guadeloupe during Martinique and Guadeloupe. Source: SIH Ifremer Blue Marlin d o l p h i n f i s h s18%ea s o n a n d o u t o f t h e s e a s o n 15%,d o2lp0hin0 f8ish. sSeaosuonr caned :o uSt oIHf th eO sebasodn,e 2b00.8 . Source: SIH Obsdeb. (“Système d’Information Halieutique”). Mathieu et al 2013

Nowadays a clear difference between the three islands species fished, represented 44% to 3be6tw%een 1o5f% tahnde 2 5t%o t(daalta FfroAm DRe ynal et al. 2013). In mately of FAD vessels are exclusive in the FAD fishing. is the FAD density present in the sea. Martinique and catches, respectively. At the sam2e0 08 tainmd e20,1 0 dino Glpuahdeilnou pef, itshhe three main species, The fish pots represent about the same proportion in yellowfin tuna, dolphin fish and blue marlin were 36% to Guadeloupe and even more, 34%, in Martinique. In Dominica have about the same number of FAD in their catches were between 9% and 4%35 %a, n33d% toy e29l%lo awnd f1i8n% ttou 9n%a r espectively of total Guadeloupe the most common strategy applied by FAD FADs landings (data from Guyader et al. 2010). Finally in fishermen is to use multipurpose fixed gears in addition to waters, 29 in Dominica (Defoe and Philbert, 2013), around Dominica we obtained in 2010 32% for the yellowfin tuna, the FAD activity (represent 31% of the FAD vessels), see 16% for the dolphin fish and 7% for the blue marlin (data Figure 7. from Fisheries Division database, 2013). See Figure 5. Faced with these results, the question is: what is the Basically, we found the same species were captured in attractiveness of FADs compared to the other activities? the three islands but in different proportions. Main species Thanks to studies carried out in Guadeloupe, we were captured are the yellowfin tuna (represented in yellow in able to note several observations. In terms of quantity Figure 5) with the highest number of catch in Guadeloupe, landed, FAD fishing and trolling line seem to be very attractive compared to other fishing activities such as pots with decreasing num b erMsa tthoiewu,a Hr.d e tt ahl.e sGoCuFtI:h6,6 s(2a0m14e) for the Page 513 dolphin fish (represented in light green in Figure 5), or gillnets (coastal activities) even if there is an important betweean n1o5%th aenrd i2m5%p o(drattaan ftro smp Reecyineasl ecta apl.t 2u0r1e3d). oInn FAmDate lay nodf FFAiDn avlelsyse ltsh aere excluvsivaer iian bthilei tFyA Df aficshtoinrg . in FAD fishing (Figure 8). In average, 2008 and 2010 in Guadeloupe, the three main species, The fish pots represent about FthAe Dsam ef isphroipnogrt ioinn dinu ces bigger gross revenue than coastal blue Marlin (represented in blue in FGiugauderleo up5e) anisd emveon rem ore, 34%, in Martinique. In yellowfin tuna, dolphin fish and blue marlin were 36% to activities (figure 9). However when you deduce the costs of 35%, 3c3a%p ttou r2e9d% iann dt h1e8 %S otou t9h% (rMespaerctivneilyq uofe )to ftaol r 20G1u0ad. eloupe the most common strategy applied by FAD FADs landings (data from Guyader et al. 2010). Finally in fishermen is to use multipurpose foixpeed rgaetairos nin (afdudietilo,n itco e, food, gear, maintenance…) the income FAD fishing started with young thfei sFhAeDr macteivnit y i(nre prtehseen t 31% of the FAD vessels), see Dominica we obtained in 2010 32% for the yellowfin tuna, for the FAD fisherman is closer but still above the income 16% foFr rtheen dcohlp hiisnl afisnhd asn,d t7h%e ifor r fthire sbtl uree msaurlitns (dianta termFisg uoref 7 .q uantities of from Fifsihsehri es lDainvdisieodn damtaboaset, 20li1k3)e. lSye e Fciogunrev 5in. ced olderF acfeids hweitrhm theesne retsou lts, thec qoumestiionng i sf: rwohmat ifsi sthhe pots and gillnets (Figure 10). Also, it is Basically, we found the same species were captured in attractiveness of FADs compared tiom thpe oorthtearn atc titvoit iepso? int out that the variability for FAD fishing practice that type of fishery (Figure 6).T hIankdse toe dst,u diFesA caDrr ied out in Guadeloupe, we were the three islands but in different proportions. Main species can put the fisherman in a dangerous economic situation in capturefdi sahrei nthge yloelolokwefidn tausna a( repprroesdenutecdt iivn ey efllioswh einr y coabmlep taor enodt et ose vceoraal sotbasle rvations. In terms of quantity landed, FAD fishing and trolling line seem to be very Figure f5i)s whiethr itehes hisguhecsht nuams befri sohf icnatgch ipno Gtusa dpelroaucptei, ced at that time by a a short period due to the cost of the associated with a with decreasing numbers toward the south, same for the attractive compared to other fishinfgis ahcitinvigti etsr siupc.h Fasi npaotlsl y, if you take in consideration the size of dolphinm faisjho r(irteypr esoefn tefdi sihn elrigmhte ngr.e enT hini sF icguhrae n5g),e ofo r fgiilslnheitns g(c oabsetahl acvtiivoitrie s) even if there is an important anotherf rimopmor toanltd seperc piees ocappltuer ecdo oun lFdA Dpo asnds iFbinlyal lyh athve e anv airmiabpilaityc tf aocnto rc ion aFsAtaDl fishingt h(Fei gucrree 8w). Iann adv etrahgee , hours fished, fish pots and gillnets may blue Marlin (represented in blue in Figure 5) is more FAD fishing induces bigger groassp preevaenru ein t hsaon mcoea sstailt uations as more interesting business than resources by exploiting them less. activities (figure 9). However when you deduce the costs of captured in the South (Martinique) for 2010. FAD and trolling line on a monetary approach (Figure 11). FAD fis hing started with young fishermen in the operation (fuel, ice, food, gear, maintenance…) the income FrenchF isilsanhdisn, gth eAir cfitrisvt rietsiuelsts iin tAermdsd oift iqounan ttitoie sF oAf D Ffoirs thhee FrAyD, Wfishheremraen is closer bFutr sotimll a btohvies t,h eo inceo mcea n easily understand why the choice of Figure 1. Number of FAD vessels from 1985 to 2009 in fish landed most likely convinced older fishermen to coming from fish pots and gillnetFs A(FiDgu rea 1c0t)i.v Aitlsyo , iat nisd the intensity of this activity are not the Effort is Allocated important to point out that the variability for FAD fishing practice that type of fishery (Figure 6). Indeed, FAD homogeneously distributed within the fleet. Non-monetary Martinique and Guadeloupe. Source: SIH Ifremer fishing lookedT ahse a pirnodcurcetiavse efi shienry cFoAmpDar edf itos hcoinasgta l doecsan pnuot th e nfieshceermssaan riinl ya dangerous economic situation in a short period due to the cost of the associated with a fisheriems seuacnh sa s afi shdineg speorttsi oprnac tiocefd atth teha t otitmhee bry af ishing activities. For incentives can also be seen to influence fishermen’s (“Système d’Information Halieutique”). majority of fishermen. This change of fishing behavior fishing trip. Finally, if you take inb ceonhsaidveiraotiro n( Gtheu ysizaed oef r et al. 2013). from oledxera pmeopplee c oiunl dt phoes siFblrye hnacveh a nis ilmapnacdts o ni nco a2s0ta1l 0, othne lcyre w2 0a%nd tahep phoruorxs if-ished, fish pots and gillnets may resources by exploiting them less. appear in some situations as more interesting business than Martinique FAD fisheries FAD and trolling line on a monetary approach (Figure 11). Fishing Activities in Addition to FAD Fishery, WhDolphinere FishFr Seasonom this, one can easily understand why the choice of the Effort is Allocated FAD activity and the intensity of this activity are not The increase in FAD fishing does not necessarily homogeneously distributed within the fleet. Non-monetary means a desertion of the other fishing activities. For incentives can also be seen to influence fishermen’s example in the French islands in 2010, only 20% approxi- behavior (Guyader et al. 2013).

Dolphin Fish Off-Season

13 FADs deployed Dolphin Fish season Dolphin Fish Off-season

Dolphin Fish 13% F i g u r2%e 4. Landing structure of the catches around FADs in Martinique and Guadeloupe during Blue Marlin 39% d o l p 42%hin fish season and o ut of the season, 2008. Source: SIH Obsdeb. Sailfish 5% 5% Figure 4. Landing structuMathieure of teth eal 2013catches around FADs in Martinique and Guadeloupe during Marlin Species 4% dolphin fish season and out of the season, 2008. Source: SIH Obsdeb. Figure 2. Number of FAD vessels from 2006 to 2011 in Figure 3. Airplane survey of FADs above Guadeloupe and Martinique and Guadeloupe; and from 2008 to 2012 in Martinique, the red spots representing the FADs. Source: Dominica. Sources: SIH and Fisheries Division of Dominica. Ifremer, 2008. CPUE

GEAR DENSITY

Figure 25. Relationship between catch of tunas in FAD associated Atlantic and Indian Ocean purse seine sets (From Kaplan et. al. 2017. Joint ICCAT, IATTC and IOTTC FAD Working Group, Madrid, Spain, April 2017) FAD density plays a role on the definition of CPUE as either an index of abundance

or an index of fishing effort interaction

CATCH CATCH PER FAD PER CATCH

NUMBER OF FADs NUMBER OF FADs Figure 26. Catch per month in Eastern Caribbean moored FAD associated fisheries on number of FADs (left panel) and catch per FAD per month on number of FADs (right panel) (Data extracted from CRFM 2015). CPUE regionally originated for the same species from different countries may reflect very different signals of abundance, several reasons that impact such statistical

Relative Contribution to Yearly Landings (by weight) 40

20 Percent Landings ofPercent

0 yellowfin blue marlin blackfin tuna dolphinfish rainbow runner little tuny skipjack

Haiti Guadeloupe Dominica Martinique

Figure 19. Relative landings of top 7 species for 4 countries participating in the moored FAD development program in the Caribbean. Estimates from Valles (2015) and Heloise et al. (2013). Conflicts among users: illegal fishing, poaching, assuming property rights that affect proper collection of fishery statistics in originated in FAD fisheries

Figure 27. Disputes registered between recreational billfish fishers in a catch-and- release mode and the forced taking over of the hooked marlin by mFAD fishers in the Dominican Republic (Source: available from news in the web). ISSUE OF UNITS OF STOCK FOR MANAGEMENT