Fishing Gear Exchange Pilot in Madras, Protected Area of Three Bays, Northeast Haiti

0 Caribbean Marine Biodiversity Program Cooperative Agreement No. AID - OAA - A14 - 00064 Fishing Gear Exchange Pilot in Madras, Protected Area of Three Bays, North e ast Haiti Submitted by Caroline Bissada, July 2019 This report is made possible by the generous support of the American people through the United States Agency for International Development (USAID). The contents are the responsibility of The Nature Conservancy and its partners (CEBSE, FOPROBIM, C - CAM, and SusGren) and do not necessarily reflect the views of USAID or the United States Government. 1 Table of Contents Introduction ................................ ................................ ................................ ................................ ..... 2 Description of fishing area ................................ ................................ ................................ ........................ 4 Landing sites ................................ ................................ ................................ ................................ ............. 4 Fishing grounds ................................ ................................ ................................ ................................ ........ 4 Travel and trap placement ................................ ................................ ................................ ........................ 5 Bait ................................ ................................ ................................ ................................ .......................... 6 Methodology ................................ ................................ ................................ ................................ ... 7 Equipment ................................ ................................ ................................ ................................ ................ 7 Trap protocol ................................ ................................ ................................ ................................ ............ 7 Seine net protocol ................................ ................................ ................................ ................................ .... 8 Monitoring schedule ................................ ................................ ................................ ................................ . 8 Student assistants ................................ ................................ ................................ ................................ ..... 9 Fishers and warden training ................................ ................................ ................................ .................... 11 Result s ................................ ................................ ................................ ................................ ........... 12 Species composition ................................ ................................ ................................ ............................... 13 Seine net fishery ................................ ................................ ................................ ................................ ............................. 13 Trap fishery ................................ ................................ ................................ ................................ ................................ ..... 14 Small - mesh traps ................................ ................................ ................................ ................................ ............................ 14 Large mesh - traps ................................ ................................ ................................ ................................ ............................ 15 GSPP vs Independent surveys ................................ ................................ ................................ ................................ ......... 21 Catch per Unit of Effort (CPUE) ................................ ................................ ................................ ................ 23 Economic value of catch ................................ ................................ ................................ ................. 24 Fisher per ception ................................ ................................ ................................ ........................... 25 Recommendations ................................ ................................ ................................ ......................... 26 Traps & Limitations ................................ ................................ ................................ ................................ . 26 Funding & Incentives ................................ ................................ ................................ .............................. 26 Capacity building ................................ ................................ ................................ ................................ .... 26 Monitoring ................................ ................................ ................................ ................................ ............. 27 Summary ................................ ................................ ................................ ................................ ....... 28 References ................................ ................................ ................................ ................................ ..... 29 Appendix I ................................ ................................ ................................ ................................ ..... 32 Appendix II ................................ ................................ ................................ ................................ .... 33 Ap pendix III ................................ ................................ ................................ ................................ ... 35 Appendix IV ................................ ................................ ................................ ................................ ... 36 Appendix V ................................ ................................ ................................ ................................ .... 38 2 Introduction Throughout Haiti, fish traps and seine nets with small mesh size are typically used by fishers who fish within the nearshore area. Fishing with this type of gear harvests large quantities of immature fish. It was clear from visual assessments of fisher’s c atch that continuing to harvest fish in this manner would soon decimate the nearshore fisheries, a resource many Haitians rely on as their sole source of subsistence. The Gear Swap Pilot Project (GSPP) was implemented under the United States Agency for International Development (USAID) funded Caribbean Marine Biodiversity Program (CMBP). The pilot was led by The Nature Conservancy (TNC) in collaboration with La Fondation pour la Protection de la Biodiversite Marine (FoProBim) and aimed to promote sustain able fishing practices, improve abundance and species diversity within the Three Bays Protected Area (PA3B) . Over a period of ten months , the GSPP fisheries consultant along w ith students from the Université d ’Etat d’Haïti , Campus Henri Christophe de Limonade work ed with fishers to replace their small mesh traps with large mesh traps . This optimized their fishery resource production by substantially increasing the size of the fish caught in the se traps while reducing juveni le catch and bycatch mortality. Seine net fishers were also approached to exchange their seine nets for large mesh traps as part of the GSPP . However, seine fishers were reluctant to participate for fear that the use of traps would not be as profitable as their seine nets. Fishers were also concerned about the loss of their investment in the seine fishery by having already purchased expensiv e seine net s . Some seine fishers said they would h ave considered participation if the pr oject bought back their nets as part o f the i ncentive pac kage . Some initial data collection on the seine fishery is presented in this report . Fishers who participated in the GSPP agreed to exchange two small mesh fish traps for one large mesh trap, provid e access to their catch for data collection , receive a weekly stipend to cover losses potentially incurred from adapt ion to fishing with new equipment as well as their ( or a family members’) participation in the CMBP’s breadfruit flour production supplementary livelihood activity . This agreement was formalised through a contract between GSPP participants and FoProBim (Appendix 1) Numerous discussions were held with fishers regarding the size of trap they preferred using (Figure 1 ) taking into consideration their varied boat sizes . Although the large mesh size of 6 cm x 5.5 cm (Figure 2 ) would remain standard, fishers expressed a need for two sizes of traps . This way, fishers with smaller boats could also participate (Fig ure 3). As such, two sizes of fish traps were provided ( small : 86 cm in length and large : 186 cm in length ) (Figure 4 ). The exchange ratio remained the same (2:1), irrespective of the size of fish trap chosen and f ishers could choose to have one size or a mix of both sizes. 3 (a) (b) Figure 1 . (a) Discussion with fishers, fisheries consultant and FoProBiM representative on the preferred fish trap size ;(b) Assessing the size of the new large traps with increased mesh size. Figure 2 . GPSS t rap mesh with increased gap size of 6.5 cm x 6.5 cm. Figure 3 . GPSS fisher in Caracol with his pirogue that can only accommodate small size traps. 4 (a) (b) Figure 4 . New fish traps with large mesh. (a) s mall traps and (b) large traps Description of fishing area Landing sites Madras and Cara c ol Bay are both surrounded by thick mangrove and sea grass beds (Figure 5 ) . The Caracol Bay landing site is located on a long stretch of beach, roughly 200 m in length . Many fishers land their catch on the e astern section of the beach close to the main road, mooring their boats slightly offshore. Other fishers land their catch at various sites along the beach. For the most part, f ishers work with specific vendors (primarily women) who wait for them to return from sea each day . Upon their return , vendors wade out in knee high water to meet the fishers. Other vendors wait on the beach under the trees near the landing sites . Some fish are processed directly on the beach whil e others are taken away to market in ice boxes . The M adras landing site is nestled d eep within the mangrove s and is a fraction of the size of Caracol Bay . There is a small broken concrete jetty that extends from an inlet within the dense mangroves. Fishers must travers e mud and muck to get to and from their landing site where some vendors sit , waiting for the boats to return . Other fishers have vendors which are located in town; t heir catch is quickly offloaded and transported via motorcycle to market. Fishing grounds The northern coast of Haiti is extremely exposed and is surrounded by a 20 km barrier and fringing reef system that extends from Bord de mer Limonade to Fort Liberté B ay. This reef system is estimated to be 1100 ha and cover approximately 10% of the shallow shelf area within the PA3B. Depths rang e from 1 m to 30 m. A deep 30 m channel separates the reef offshore of the Caracol Bay. Fishers have been observed using a number of distinct coral reef zones, including reef crests, backreefs, reef walls, patch reefs, and spur - and - grove reef habitats. Spu r - and - groove reefs are found directly in front of Caracol Bay. On the backreef of the spur and groove lies a descending slope of complex r u gose hard corals interspersed by sand. Reefs are subject to effluent from both the Grande Rivière du Nord and the Cap Haïtien rivers. 5 F igure 5 . Map showing the Madras and Caracol landing sites (blue dots) as well as the trap fishing sites (yellow flags) located on the barrier reef. Dense mangroves can be seen surrounding the landing sites. A total of 51 coral species, 43 octocoral species, and over 140 species of sponge have been identified within the PA3 B reef habitats (Kramer et al 2016) . The most frequent corals are mustard hill coral ( Porites astreoides ) , lettuce coral ( Agaricia agaricites ) , brain coral ( Pseudodiploria strigose ) , and star coral ( Montastraea cavernosa ) (Kramer et al. 2016) . Also found within the shallow reefs of the PA3B is the cup coral ( Cladopsammia manuelensis ) which typically occurs as part of deep reefs (>50 m) systems in the Straits of Florida and the Gulf of Mexico (Kramer et al. 2016) . Caracol Bay has an estimated 4,030 ha of mangrove with an extensive network of channels occurring throughout. Mangrove stands include the black mangrove ( Avicennia germinan s ) , red mangrove ( Rhizophora mangle ) , white mangrove ( Laguncularia racemose ) and buttonwood ( Conocarpus erectus ) . The seaward border and hammock mangroves experience regular tidal action, wh ilst the landward basin mangroves only experience periodic flooding. Adjacent to the mangrove stands, are highly productive seagrass meadows . Prior to the GSPP fishing took place adjacent to the thick mangrove forests in the Caracol B ay and Madras bay. Now, with the large mesh traps fishing takes place in deeper water near coral reef habitats. Travel and trap placement The closest coral reef is the barrier reef located approximately six km offshore from both the Caracol Bay and Madras landing sites (Figure s 5 ). One fishing trip takes an average of seven hours. Fishers have small, open, wooden pirogue boats with makeshift sails and bamboo masts (Figure 7 ). Fishers primarily fish in groups of two to four persons , though th e re may be as many as six per boat . Fishers leave their landing sites in the dark between 4 - 6 am and row out to the barrier reef. This takes roughly three hours. Fishers then haul and re - set their traps (Figure 6) . GSPP fisher s were provided with one cooler (igloo) each and fish were placed directly in to the cooler after hauling. Ice is a scarce commodity and comes in the form of a giant block. Most fishers try to obtain ice prior to setting out on their fishing trips to help preserve the freshness and integrity of the fish, though it is not always possible. The amount of time 6 spent at sea then becomes crucial , as spoilage can occur , even when the harvest in placed in coolers when there is no ice available, as fishers often return to their landing sites by mid - day. Fishers sail back to their landing sites , which require s less manual effort but can be still be time consuming if the wind is light (Figure 7 ) . Large mesh t raps we re placed along the edge of the barrier reef on both the inner patch reef sections, the backreef as well as along the slope towards the channel . (a) (b) (c) Figure 6 . The shallow reef flat “ at oll”. (a) Boats can be seen in the channel (La Pase), the edge of the barrier reef is seen in the distance (right) , (b) a Caracol fisher walking on the barrier reef to place small mesh trap s (right), (c) fisher navigates the boat on the shallow reef by pushing with an oar (bottom) . Figure 7 . Fishers use wooden pirogues, bamboo oars and masts with sails sown together from pieces of material. Bait Fishers would use seasonal fruit such as mangoes as bai t in their traps. However, once they switched to using large mesh traps the bait proved less efficient. This was likely due to the change in the target species from herbivorous species found in the mangrove and seagrass meadows to pi s civorous species found on coral reefs and deep slopes. Fishers were counselled by the fisheries consultant t o affix small to medium sized shiny objects to the bottom of their traps so that the reflective object would resemble prey fish with its movement in the water and attract piscivorous fish into the traps. Fishers reported a higher rate of success with this new bait type. 7 Methodology Equipment - Three 5 - gallon buckets (5 or 6 buckets for seine net protocol) - 1 x pulley scale 128 kg x 2 kg - 1 x hanging hook fishing scale or flat kitchen type scale - 1x fish measuring board - Caliper or small measuring tape - Small bucket to use in conjunction with hanging hook scale - Slates made from cute Perspex, sand ed with sand paper - Extra sand paper - Sharpie type permanent marker - Pencils - String or flagging tape to hold pencil to slate - Sponge, soap to wash slates - Camera (on phone) to record picture of slate prior to washing for permanent record of data T rap protocol Small mesh traps are the standard trap used in the Caracol and Madras bays. The size of the mesh opening is 3.8 cm x 3.8 cm. For each day fishing, the catch was weigh ed to get the total biomass , then data was collected on the species composition, individual fish length and individual fish weight. W eight (kg) of the empty large bucket using the pulley scale (128 kg x 2 kg) 1. The entire catch is placed in the bucket and weighed (kg) using the pul ley scale 2. For each individual fish the following was recorded : a. The fish species ( at minimum the family level ) . b. The total length (cm) using a measuring board . c. The weight (kg) for each fish using a hanging hook fishing scale . 3. For lobster the following was recorded : a. The carapace length (cm) . b. The weight (kg) for lobster is weighed using a hanging hook fishing scale . c. T he presence or absence of eggs was recorded . 4. S tart and end time of the fishing trip (e.g. 5 am to 10 am) wa s recorded . 5. S oak time (number of days submerged) of the traps is recorded . 6. N umber of fishers on each trip (e.g. 2 fishers) . 7. A brief description from the fisher of the areas where the traps were placed (e.g. local name of the area, substrate type, depth) wa s recorded . 8. Gear type and size (l arge or small ) . weight of bucket with catch (kg) – weight of empty bucket (kg) = total weight of catch (kg) 8 Seine net protocol The number of fishers that operate one seine net varies between six and eleven depending on the size of net. A typical seine net is approximately 300 m wide with a mesh size of roughly 1 cm in diameter. Two boats are needed to place and haul the net and f i shers on both boats haul continuously for over 5 hours. Fishing trips can last up to 11 hours, sometimes coming in with no catch at all. The c omposition of the catch sample for seine nets was similar to that of the traps . However, the seine nets caught a much larger volume of small fish. Data collected from seine nets was exceptionally time consuming because the fish caught were very small and their individual weight often did not register on the scale. One sample bucket c ould contain over 500 individual fish. Data on catch composition were collected from the catch of five seine nets on five separate days between June 5 th and July 5 th , 2018 . at the family level and species level when possible. Total length (cm) was recorded for each fish in th e sample bucket and the weight was obtained when possible. Total biomass was captured by weighing the entire catch as well as the number of fishers and the total fishing trip time. Once the catch was landed it was evenly distributed amongst buckets. One bu cket was chosen at random as a subsample to determine species identification and size. Sampling occurred as follows: 1. All sea snakes, crustaceans, debris and seagrass were removed. 2. T he weight (kg) of the empty buckets was taken. 3. T he entire catch in all the buckets w ere e qually and randomly distribute d. 4. The weight of each bucket filled with fish was taken and the weight of each empty bucket subtracted ( weight of bucket with catch – weight of empty bucket = weight of catch of bucket ) . All weights of buckets containing fish were then added to obtain the total weight of the catch. 5. O ne bucket was randomly selected to sub sample the catch in detail 6. For each fish in the subsample : a. The fish was identified to the species level, at minimum the family level b. The total length was recorded on the measuring board c. The mass of the fish was weighed and recorded With respect to Caribbean spiny Lobster measurements , the size of the lobster’s carapace was recorded using a caliper or measuring tape. The front of the measuring gauge was place on the front of the carapace in between the eyes at the forward edge between the rostral horns and the gauge was extended back along the middle to the rear edge of the carapace. Monitoring s chedule Monitoring occurred every time a fisher hauled traps from June 30 th , 2018 to August 29 th , 2018 , then again from October 20 th ,2018 to November 24 th , 2018. For each fishing event, d ata was collected on species composition, fish length, the weight of individual fish , the weight of the total catch trap size, weight b1 + weight b2 + weight b3 + weight b4 = total weight(kg) 9 number of traps placed, number of traps hauled, the duration of the fishing trip, the number of fishers and t h e fishing site . St udent assistants Education and partnership with the local university was a major component of the GSPP. The pilot project facilitated nine graduating students from the Université d’Etat d’Haïti, Campus Henri Christophe de Limonade to improve their capacity in fisheries and coastal ecology. Though there i s no N atural R esource M anagement or M arine B iology stream, the G eography and A gronomy majors are required to understand the ecology and biogeography of Haiti’s coastal systems as well as the socio - economic issues affecting the people of Haiti. Students were also required to write a final research report as part of their graduat ion requirement. S tudents were given the opportunity to join the fisheries biologist and gain on - site ex perience receiv ing valuable training in field data collection methods , including learning to identify a variety of reef, mangrove and sea grass species and collect a range of metrics such as total length, fork length and weight . Students were also allowed to use extra data collected for their own research and the time spent with fishers to conduct their own additional interviews for their individual projects (Figure 8 ) . A stipend was given to students to cover expenses such as daily travel to and from the landing sites and food. Students in groups of four rotated field days with the fisheries biologist. Many of the students also had an opportunity to experience the reef, mangrove and sea grass ecosyste ms by participating in excursions with the fisheries consultant and the fishers. A database was created to house the large amounts of data collected on species composition, catch per unit effort and other allometric variables such as total length and wei ght . The students were trained in both data entry and data management of the large spreadsheets and were a part of the ongoing socio - economic discussions with fishers (Figure 8 ) . With the help o f the students , a database of fish species was compiled with both English and Haitian creole names to facilitate the exchange of information and ensure accuracy of the data being collected. Some species have their own specific local name whilst others were grouped as families and a generic family name used (Ap pendix I I , Table I ). Students received certificates outlining their achievements which can be used by them to certify that they have undergone this type of training (App endix II I ) . 10 Figure 8 . Training of students in field data collection, species identifcat i on and data management . 11 Fishers and warden training Fishers, wardens, NGO personnel as well as other stakeholder s were trained in in the field in fisheries management data collection protocols ( F igure 9 ). Groups were kept small (no more than four persons ) to be able to clearly exchange information and to be able to offer valuable hands - on practice without causing delays to fishers. Presentations on ecosystem processes, key species and invasive species were given to a variety of stakeholders including fishers, fisheries co - operative members, students, wardens and fish vendors from different areas of the PA3B. A large number of fisheries stakeholders from Caracol, Limonade and Fort Li berte expressed interest in learning data collection methods but did not have the opportunity to participate. Ongoing capacity building in this area is recommended in the future. Figure 9 . Fishers and wardens acquire training in species identification, an d allometric data collection of the fisheries resources. 12 R esults The GSPP was successful in promoting sustainable fishing practices and livelihood s to fishers by reducing the impact on their juveniles and providing a higher income to fishers , facilitating them fishing less to earn more money . Figure 10 . Typical catch from the Madras landing site from small mesh traps. Figure 1 1 . Typical catch from the Madras landing site from large mesh traps. 13 Species composition Seine net fishery S eine net data was collected from five seine nets early in the data collection period however, it was discontinued when no seine fishers officially signed up to exchange their nets for traps. Focus then shifted to compare the catch composition between the small mesh and large mesh size traps. In one 10 kg sample bucket, 355 fish were identified and measured (Table 1). An extra 100 fish were not measured from the sample bucket as the fishers were pressed for time to go to market. Of the 355 fish sampled, 265 were molly millers of the Blenniidae family (Figure 12 ) . Fish size ranged from 2 to 10 cm with the mean fish length being 6.44 cm. Molly miller s inhabit sand bottoms and mudflat areas and grow to a maximum 12 cm in length. Parrotfish were the second main family of fish caught (n =71) with seine nets . Most parrotfish captured were in the juvenile phase making it difficult to identify to the species level. Parrotfish ranged between 3.5 cm and 8.2 cm in length, with the mean being 5.57 cm. The average weight of fish sampled in the bucket was 0.03kg. Tab le 1. Sample species composition by family or species caught in five a standard seine net between June 5 th and July 5 th 2018 . The table provides information on the mean fish total length (TL), the smallest fish’s total length (Min TL) and the longest fish t otal length (Max TL) . Family / Species Mean TL cm Min TL cm Max TL cm N umber of fish Filefish 7.05 7.00 7.10 2 Flagfin mojarra 6.15 5.80 6.50 2 Groupers 5.00 5.00 5.00 2 Grunt 6.75 5.00 8.30 5 Jack misc. 7.30 5.90 8.70 2 Molly miller 11.22 5.00 24.90 265 Snappers 7.80 7.80 7.80 1 Parrotfish misc . 6.44 2.00 10.00 71 Blue tang 5.57 3.50 8.20 3 (a) (b) Figure 1 2 . (A) Basket of catch from seine net ; (b) handful of Blenniidae species from seine net catch . 14 Trap fishery Between July and November 2018 , small mesh traps were hauled and sampled 36 times. Each fisher usually sets and hauls between 5 and 25 traps per fishing trip. D ata on catch composition including length (cm) and weight metrics (kg) were collected for all fi sh and lobster caught in the small and large mesh traps at the Madras and Caracol landing sites . Three thousand eight hundred and seven (3807) fish were identified and measured. Small - mesh traps Traps with small mesh openings are regularly placed in nu rsery grounds such as mangrove forests and seagrass beds as well as on the fringing reef itself . As a result, juvenile reef and mudflat species we re primarily caught . Table 2 shows the diversity of fish caught in in the standard small mesh traps (grouped by family). Appendix I I shows the breakdown of the fish family into species. Table 2. Diversity of fish grouped by family caught in the standard small - mesh traps. Important e cological and/or economical species are highlighted in grey. The table provides information on the mean fish total length (TL), the smallest fish’s total length (Min TL) and the longest fish total length (Max TL); as well as the mean mass/weight of all fis h (kg), weight of the smallest fish (Min weight) and the mass/weight of the heaviest fish (Max weight). Family Mean TL (cm) Min TL (cm) Max TL (cm) Mean weight (kg) Min weight (kg) Max weight (kg) N umber caught Barracuda 28.82 21.40 38.50 0.15 0.04 0.31 11 Bonefish 21.50 21.50 21.50 0.06 0.06 0.06 1 Butterflyfish 6.67 5.50 7.50 0.01 0.00 0.02 7 Cichlids 18.11 13.40 22.00 0.10 0.04 0.17 11 Damselfish 12.86 9.50 16.50 0.06 0.03 0.09 16 Drum 26.30 26.30 26.30 0.29 0.29 0.29 1 Goatfish 15.87 10.30 19.00 0.05 0.02 0.08 11 Grouper 16.59 11.10 28.60 0.10 0.03 0.42 63 Grunt 12.98 10.00 23.00 0.04 0.01 0.20 124 Hamlet 10.80 10.70 10.90 0.03 0.03 0.03 2 Jacks 21.57 11.40 24.20 0.13 0.02 0.19 11 Mojarra 12.50 9.00 17.30 0.05 0.00 0.63 41 Mullet 12.92 7.84 18.00 0.44 0.07 0.80 2 Needlefish 36.15 33.90 38.40 0.08 0.06 0.09 2 Parrotfish 14.01 8.20 26.50 0.07 0.00 0.55 2707 Porg y 15.89 0.60 30.40 0.10 0.02 0.80 111 Scorpionfish 19.50 19.50 19.50 0.19 0.19 0.19 1 Snapper 15.08 7.70 28.00 0.09 0.02 0.90 90 Snooks 22.14 18.50 28.50 0.09 0.05 0.19 8 Spadefish 19.20 19.10 19.30 0.29 0.29 0.30 2 Squirrelfish 14.72 9.30 18.50 0.04 0.00 0.07 96 Surgeonfish 12.90 7.50 20.00 0.06 0.00 0.29 47 Trumpetfish 45.00 45.00 45.00 0.43 0.43 0.43 1 Wrasse 16.70 12.50 26.70 0.08 0.00 0.28 44 Total 14.25 0.60 45.00 0.07 0.00 0.90 3410 15 There were 2,707 fish of the parrotfish family found in the small mesh traps sampled . Parrotfish were found in every fishers ’ catch each time the catch was sampled . Parrotfish accounted for 79% of the total catch in the small mesh traps. The number of parrotfish caught per fisher on a sampling day varied between 1 and 347. On average fishers caught 9 parrotfish per small mesh trap, with a r ange of 1 to 19 fish per trap. Juvenile parrotfish ranged from 8.30 cm to 26.50 cm in length (mean total length of 14.01 cm) and had a mean weight of only 0.07 kg ( Table 2 ) . O nly 11% of the sampl ed parrotfish from the small mesh traps were found to be sexually mature ( Table 4 ) , if a general L 50 length at sexual maturity is used . Other ecologically and economically important species caught in small mesh traps were porgies (n= 111), snappers (n= 9 0) and groupers (n= 63) , see Table 1. The seabream porgy matures at small lengths, thus 47 % of the seabreams sampled in the small traps could have been mature individuals (see Table 4) . On the other hand, 100% of all snappers caught in small traps were l ikely to have been sexually immature. The 11 G reat B arracuda s whose sizes ranged from 21.4 to 38.5 cm in length (see Table 4) that were caught were also likely to be sexually immature . R eef species such as grunts, squirrelfishes and wrasses were harvested in abundance (n= 124, n= 96 and n= 44) . Results show that all wrasses captured were likely to have been sexually mature. However, only 55 % of the squirrelfish and 5% of the grunts caught were sexually matu re (see Table 4). The use of small - mesh traps also caught non - targeted ornamental fish as bycatch such as Butterflyfish (n=7), Damselfish (n= 16), Drums (n=1) which were discarded by some fishers. Large mesh - traps Traps with the large mesh (6 cm x 5.5 cm) are designed to allow small and juvenile fish species to escape and the bigger, already reproductively active, adult individuals to be retained . The change in mesh size and subsequent change in fishing grounds was meant to relieve pressure off juve niles, allowing them to grow to maturity, spawn and hopefully maintain the fish stocks. Table 3 shows the diversity of fish (grouped by family) caught in CMBP GSPP traps that were designed with larger mesh openings. Appendix I I shows the breakdown of the f ish family into species. Although some parrotfish are still caught in the traps there has been a significant decline from 79% of total catch in small mesh traps to 16% of the total catch in large mesh traps. Individuals caught in the new gear are on average 2 cm longer than the parrotfish caught in the small mesh traps . Even the smallest individual caught in the large mesh trap is 3 cm longer than the smallest fish caught in the small - mesh traps. Approximately 30% of the parrotfish caught in the large trap had reached lengths at which sexual maturation occurs. While the large - mesh traps did not guarantee that on ly sexually mature parrotfish would be caught, the fish in the large - mesh traps were larger, and depending on the specific species of parrotfish, some could potentially be mature (see Table 4) . Like many reef species, parrotfish are protogynous hermaphr odites, changing from female to male once they reach a certain size. Females have been found to change to males at fork lengths ranging between 26.2 - 47.3 cm ( Hawkins and Callum 2003 ) . Overharvesting individuals below the change threshold could drasticall y alter the female - male ratios needed to sustain stocks. The large - mesh traps caught larger individuals and fewer of them potentially increasing the likelihood that the ratios of females to males would be at an optimal level ( Hawkins and Callum 2003 ) . 16 P orgies were the most harvested species in the large traps accounting for 35% of the total catch (n = 89), then parrotfish, 16.6% (n=44), then snappers, 13.5 % (n= 36). Of the porgies, 86% of the seabreams were likely to have reached sexual maturation (L 50 = 8 cm) as were 66% of lane snapper s (L 50 = 20.6cm) (see Table 4). The Lane snapper is listed on the IUCN list as near threate ne d due to a decrease in the number of mature individuals across its range. All Caribbean spiny lobster (n=11) harvested in the l arge traps were mature and had carapace lengths ranging between 5.5 and 9. 4 cm in length (see Table 4 ) . Additionally, none were egg bearing at the time of capture. The large - mesh traps eliminated the by - catch of ornamental fish species entirely. Large - mesh traps a re ineffective ( mesh size results in no catch) when place d on the shallow spur and groove section s of the reef . The use of large - mesh traps has stopped fishers from placing traps directly on t he reef and accordingly has reduced the walking on the reef by fishers . Table 3. Diversity of fish grouped by family caught in GSPP traps with large - mesh. Important ecological and/or economical species are highlighted in grey. The table provides information on the mean fish total length (TL), the smallest fish total length (Min TL) and the longest fish total length (Max TL) and c arapa ce length for spiny lobster*; as well as the mean weight (kg) of all fish, weight of the smallest fish (Min weight) and the mass/ weight of the heaviest fish (Max weight). Family Mean TL (cm) Min TL (cm) Max TL (cm) Mean weight (kg) Min weight (kg) Max weight (kg) N umber caught Barracuda 35.60 35.60 35.60 0.24 0.24 0.24 1 Billfish 29.70 29.70 29.70 0.10 0.10 0.10 1 Boxfish 27.82 20.00 38.50 0.48 0.18 0.89 5 Drums 23.35 22.70 24.00 0.18 0.16 0.20 2 Goatfish 25.63 23.60 29.00 0.25 0.16 0.32 3 Grouper 25.54 20.70 29.50 0.28 0.15 0.46 7 Grunt 18.41 12.00 30.70 0.11 0.02 0.39 14 Jacks 21.03 11.20 29.50 0.13 0.02 0.26 16 Mojarra 13.12 8.80 27.10 0.05 0.01 0.20 22 Panulirus argus 28.74 5.50* 90.40 0.40 0.15 0.67 11 Parrotfish 16.12 11.20 30.30 0.09 0.02 0.32 44 Porg y 17.06 11.60 32.20 0.10 0.03 0.32 89 Snapper 18.06 12.20 32.60 0.16 0.03 0.75 36 Spadefish 13.20 10.20 18.20 0.10 0.04 0.20 3 Squirrelfish 21.05 16.50 25.60 0.10 0.03 0.16 2 Surgeonfish 12.35 10.10 18.20 0.05 0.03 0.12 6 Scorpionfish 31.50 30.00 33.00 0.38 0.33 0.44 2 Pufferfish 15.00 15.00 15.00 0.07 0.07 0.07 1 Grand Total 18.18 5.50 90.40 0.13 0.01 0.89 265 Other economically important fish harvested in the large - mesh traps were groupers. Of the grouper and sea bass family caught, two were Nassau grouper or “barré” in Haitian creole (TL = 17 and 13.5 cm) and one was the red grouper (TL = 29.5). Though the Nassau groupers had not reached sexual matur ity, the red grouper was likely mature ( L 50 = 19.9 - 19 cm). Other species of grouper such as the yellowfin grouper were also being caught in the large mesh traps ( see Figure 1 3 ) th ough they were not present in traps during the sampling period. As with the parrotfish, groupers are protogynous hermaphrodites thus, it is 17 important for individuals to reach large enough sizes in order to maintain female - male ratios and ensure continuity of local populations. The Nassau grouper is listed as threatened across its biological range under the Endangered Species Act and is included in Annex III of the SPAW protocol which states that exploitation is allowed but should be regulated to ensure and maintain the populatio n at an optimal level. Similarly, the yellowfin grouper (Figure 1 2), the mutton snapper which was captured in both small and large traps and the lane snapper are listed as near threatened on the IUCN red list. Fragmented populations and a decline in mature individuals could pose a threat to stocks in the near future. Currently there are no regulations protecting these important species in Haiti. As such harvesting these individuals is legal. Fishers could benefit from trainings on the identification of spe cies at risk and on maturation metrics such as length at maturity. Being able to identify key species and return them to the water (with minimum harm to fish) could help protect their stocks going forward. At the moment the large - mesh trap fishery is limit ed in scope, size and vessel power however c ombined with other fishing practices such as spear fishing, hookah and the use of small - mesh seine nets in the same habitat adds a cumulative and compounded impact; fishers do apply constant pressure to the fishe ry as it is their main source of subsistence. 18 Table 4. Percentage of sexual mature species sampled in traps. Sexual maturation is given as the size at which 50% of the sampled species have reached sexual maturation (L 50 ). If the L 50 was different for males and females the female range was used. When studies obtained varying results, a range was used. The r ange of fish size (smallest and longest) is noted as the to tal length in cm. The number of individuals of each species found in both the small and large mesh nets is noted as well as o verall sample size of each species . Species Size at L 50 (cm) TL length range (cm) Number of mature individuals Sample size % mature individuals in the catch small mesh large mesh small mesh large mesh small mesh large mesh small mesh large mesh Atlantic Spadefish 12 19.1 - 19.3 10.2 - 18.2 2 1 2 3 100 33 Banded Butterflyfish 12 5.5 - 7.5 - 0 0 6 0 0 - Bar Jack 31 11.4 - 20.5 11.2 - 29.5 0 0 3 14 0 0 Blue tang 15 7.5 - 12.0 - 0 0 6 0 0 - Bluehead wrasse - 18 - - 0 1 0 0 0 Bandtail puffer - - 15 - - 0 1 - - Bonefish 40 21.5 - 0 0 1 0 0 - Butterflyfi - 6.5 - - 0 1 0 - - Caribbean spiny lobster* 5 - 8 - 5.5 - 9.4 - 11 0 11 - 100 Coney 18.5 11.1 - 22.5 20.7 - 27.0 12 4 60 4 20 100 Flagfin mojarra - 10.0 - 17.3 14.1 - 27.1 - - 22 3 - - French grunt 15 - 17 10.0 - 20.7 13.2 - 29.2 6 3 101 3 6 100 Goatfish misc. - 10.3 0 - 0 1 0 - - Gray s na pper 33 13.0 - 26. 1 12.2 - 26.5 0 0 6 5 0 0 Great Barracuda 71.0 - 98.5 21.4 - 38.5 35.5 0 0 11 1 0 0 Grunts misc. - 13.3 - 23 14.1 - 21.0 - - 13 6 - - Honeycomb cowfish 22.2 0 20.0 - 38.5 - 4 0 5 - 80 Houndfish 51.7 33.9 - 38. 4 - 0 - 2 0 0 - Indigo hamlet - 10.8 - 10.9 0 - - 2 0 - - Irish pompano - 10.8 - 16.0 9.0 - 17.9 - - 8 11 - - Jacks misc. - 22.5 - 24.2 19.4 - 20.0 - - 8 2 - - Lane s na pper 20.6 28 15.6 - 30. 5 1 4 1 6 100 6 7 Longspine Squirrelfish 13.5 11.0 - 18.5 25.6 33 1 85 1 3 9 100 Marlin 16.4 - 29.7 - 1 0 1 - 100 Majorras misc - 9 - 14.4 - - - 10 0 - - 19 Mutton snapper 45 - 47 12.8 - 18.3 12.7 - 25.3 0 0 11 30 0 0 Nassau grouper 48 13.5 - 17 24.4 - 28.7 0 0 2 2 0 0 Oceansurgeonfish 15.1 7.8 - 20.0 10.1 - 18.2 10 1 34 6 29 1 7 Parrotfish spp. 17 - 23. 8.2 - 26.5 11.2 - 21. 5 252 12 2291 40 11 30 Princess parrotfish 20.1 13.2 20.3 0 1 1 1 0.00 100 Puddingwife - 14.2 - 26.7 - - - 15 0 - - Lionfish 18.9 - 19.0 30 - 33 - 2 - 2 - 100 Red grouper 18.9 - 19 - 29.5 - 2 0 2 - 1 00 Red hind - 28.6 - - - 1 0 - - Redband parrotfish - 9.0 - 26 .1 - - - 130 47 - - Sailors choice - 12.0 - 18.7 12.0 - 18.7 - - 4 4 - - Saucereye porgy - 20.5 13.8 - 20.6 - - 1 6 - - Schoolmaster 25 11.0 - 19.0 14.2 - 32.6 0 1 47 4 0 25 Sea bream 8 9.0 - 30.4 11.6 - 32.2 52 63 110 73 47.27 86 Sergeant major - 9.5 - 16.5 - - - 15 14 - - Silver jenny 11 14.5 8.8 - 13.8 1 1 1 8 100 12.5 Slippery dick 18.5 15 - 0 - 1 0 0 - Smallmouth grunt - 7.7 - 20.8 - - - 4 0 - - Snapper misc. - 10.5 - 18.0 - - 20 0 0 - Snook 42 18.5 - 28.5 - 0 - 8 - - - Spanish grunt 15 - 17 11.5 - 14.2 - 0 - 10 - - - Spotted drum - 22.7 - 24.0 - - - 2 - - Spotted goatfish - 15.5 - 19.0 24.3 - - 4 1 0 - Spotted scorpionfish - 19.5 - - - 1 0 - - Squirrelfishes misc. 14.5 9.3 - 17.3 16.5 6 1 11 1 5 2 100 Stoplight parrotfish 16.3 9.6 - 23.5 16.3 - 30.3 90 3 283 3 3 2 100 Striped croaker - 26.3 - - - 1 0 - - Striped mullet - 7.84 - - - 1 0 - - Surgeonfish 15.5 10.9 - 16 - 2 - 7 0 2 9 - Tilapia - 13.4 - 22.0 - - - 0 - - Trumpetfish - 45 - - 11 0 0 - White grunt - - 30.7 - - 1 1 - - White mullet - 18 - - - 1 0 - - Wrasse misc. - 12.5 - 26.5 - - - 26 0 - - Yellow goatfish 11 15.0 - 16.7 23.6 - 29.0 5 2 6 2 83 100 Yellowtail snapper 19.7 - 26.0 27 26.6 1 1 1 1 100 100 Yellowtail damselfish - 10.8 - - - 1 0 - - 20 (a) (b) (c) (d) (e) (f) (g) Figure 1 3 . (a - f ) E xamples of the large mature fish caught in the large mesh traps Verne Pierre . (a and b) Madras GSPP fisher showing his catch from the large - mesh traps . (g) C . Bissada project fisheries biologist and University of Limonade students having an informal meeting with Caraol fishers Georges Bertteau and Florvil Emilien (standing) . Fishers regularly catch Caribbean spiny lobster in their large - mesh traps. Although finding lobster in the large - mesh traps was noted from the onset of the p ilot , data collection with metrics did not begin until the data collection period in October 2018. Therefore , catch data does not reflect the true number of lobsters being caught in the traps. As lobster represents an economically beneficial fishery to fishers, continued data collection would prove valuable in identifying the pressure being applied to t his fishery. Calipers to accurately measure carapace length were brought in and students and fishers were trained in their use (Figure 14) . Discussions with fishers were held on assessing the presence and absence of eggs and fishers agreed of their own vol ition to return lobsters found with eggs to the sea. No egg - bearing lobsters were observed during the data collection periods. Eleven, lobster s were 21 caught within the traps and carapace length s were between 5.5 cm and 10.1 cm (mean 8.3 cm), with a mean wei ght of 0.4 kg. Figure 1 4 . Fisheries consultant demonstrating to the students how to correctly measure spiny lobster. GSPP vs Independent surveys Independent fish surveys and interviews with fishers at landing sites were conducted in the PA3B by the Atlantic and Gulf Rapid Reef Assessment (AGRRA) team and James Miller and listed in H enwood et al 2017 . The extensive list of marine fish species listed in Henwood et al 2017 indicated a near complete absence of typical la rge predators such as groupers. Similarly, the CMBP conducted AGGRA surveys of the reef ’ s biophysical conditions across the seascape in the PA3B and indicated that “carnivorous fishes are scarce”. Their surveys revealed 0 b arracuda, 0 c hubs, 0 - 3 groupers, 0 - 6 jacks , 0 porgies and 0 snapper. However, the data collected for the GSPP in the large mesh traps shows that jacks, porgies and snappers and mojarras are routinely caught. In only a few weeks of sampling 16 jacks , 89 porgies, 36 snapper, 22 m ojarras and 7 groupers were observed . The small mesh traps also saw barracuda caught , as well as the prior listed species. Observation from ongoing trap fishers using the large mesh traps continue d to show the presence of large predators. It is possible that the sampling depth (mean depth 7m in 2018 ) used by the AG R RA protocol does not represent the true picture of the species composition. Also, surveys conducted by divers may scare large predators away. Continued monitoring of the catch composition of the trap fishery as well as the spearfish and hookah fisheries would provide more insight. The presence of these top predators in the catch composition of the fishers within the PA3B , highlights not only their presence , but the importance of the habitat for management . A sustained monitoring program should therefore be a priority. T he independent surveys by the AGGRA team refer to the Coney ( C ephalopholis fulva ) filling the space held by the top predator s. While there is not enough data in this s tudy to address that specifically, their survey showed far fewer predators than were observed in the catch compos i tion of the GSPP fishers . 22 Other species present in the catch composition of the trap fishery and not observed during the independent surveys are listed in Table 5 below. Table 5. Additional marine fish species found within the PA3B not listed in the Protected Area of Managed Natural Resources of the PA3B Management Plan by Henwood et al 2017. Scientific name Common name Sphoeroides spengleri Bandtail puffer Albula vulpes Bonefish Eucinostomus malanopterus Flagfin mojarra Acanthostracion polygonius Honeycomb cowfish Tylosurus crocodilus Houndfish Diapterus auratus Irish pompano Caranx hippos Crevalle jacks Istiophoridae Marlin Epinephelus striatus Nassau grouper Pterois volitans Lionfish Epinephelus morio Red grouper Eucinostomus gula Silver jenny Centropomus undecimalis Snook Haemulon macrostomum Spanish grunt Parupeneus spilurus Spotted goatfish Scorpaena mystes Spotted scorpionfish Bairdiella sanctaeluciae Striped croaker Mugil cephalus Striped mullet Oreochromis niloticus Tilapia Aulostomus maculatus Trumpetfish Mugil curema White mullet Mulloidichthys flavolineatus Yellow goatfish Chrysiptera parasema Yellowtail damselfish 23 Catch per Unit of Effort (CPUE) Catch is a function of the biomass of fish in the water as well as the fishing effort used to harvest them (C = B · qE where q is a catchability coefficient). Often, catch can simply be divided by effort to get CPUE with the idea that C / E = B · q. Effort ho wever can often vary by factors such as fisher , area, or size/mesh of a trap . Catch therefore is a function of the different components that make up effort C = Traps + Soak Time + Fisher (dropping the B · q). Linear models can then be used to standardize e ffort across the different factors. Small mesh and large mesh data sheets were combined and a categorical variable denoting mesh size was added. The number of traps was considered the unit of effort that was then influenced by the mesh size, the soak time and the different fishers. Soak time was included as an offset meaning that the value w as simply added into the intercept term in the linear model instead of having an estimated parameter. Including soak time as an offset recognizes that soak time is a part of the effort. The CPUE was standardized with a mixed - effects linear model in the fre eware software R with package lme4. The mixed - effects enable the modelling of the individual fishers (called the random effects) while also producing a single population level estimate representing the expected value across all fishers (fixed effects). The response variable was CPUE, the total catch of a fisher (f) in a day (t) divided by the total number of small or large mesh traps hauled that day. Ln (CPUE) f,t = (β 0 +b 0 )+β 1 Mesh Size f,t +offset[Ln(Soak Time)] f,t +ε (1) In the model, β 0 is the intercept and b 0 ar e the deviations from the in tercept to account for the impact of t he different fishers. β 0 is the expected value across all fishers (fixed effect) and b 0 are the random effects for each individual fisher. To improve model diagnostics and produce a good fit to the data, the response variable was the natural log of CPUE an d the natural log of soak time was used. The model fit the data and had reasonable diagnostics. There was a lot of spread in the data resulting in a fit that was good, but not particularly tight. Soak time did not end up being an essential component of the model. Including it or removing it made relatively little difference to the overall fit of the model. This is likely because the majority of the data had a soak time of two days and the CPUE from a soak time of two days ranged widely. There was also no t a clear relationship between catch and soak time. Fishers caught lots of fish with long and short soak times and fishers also caught few fish with long and short soak times. Soak time may be important, particularly with a larger data set and thus it was included in the final model, but it had limited influence with the current data. With the estimated parameters, the average CPUE for a small mesh trap and a large mesh trap assuming