Assessing potential protection effects on commercial fish species in a Cuban MPA

Brbr Hort e Cost,Jorge Angulo-Vldés,Jorge M.S.Gon?lves,Pedro Brros

aCenter of Marine Sciences,CCMAR,University of Algarve,Campus de Gambelas,Faro,8005-139,Portugal

bCentro de Investigaciones Marinas,Universidad de La Habana,Calle 16 No.114,Paya,La Habana,CP 11300,Cuba

cSchool of Natural Resources and Environment,University of Florida,United States of America,103 Black Hall,Gainesville,FL,32611,USA

dCentre for Marine and Environmental Research(CIMA),University of Algarve,Campus de Gambelas,Faro,8005-139,Portugal

eFood and Agriculture Organization(FAO),Fisheries and Aquaculture Department Marine and Inland Fisheries Branch(FIRF),Rome,00153,Italy

ARTICLEINFO

Keywords:

Cuba

Marine protected areas

Commercial fish species

Spawning aggregations

Artisanal fisheries

Coral reefs

ABSTRACT

Cuba has been leading marine protected area(MPA)designation in the Caribbean region to ensure conservation of its valuable marine ecosystems.Yet,an efficient monitoring program for MPAs is still to be implemented and will benefit from pre-existing information.The highly regulated MPA of Punta Francés National Park is one of the first Cuban MPAs and was established mainly to protect species and ecosystems for tourism purposes.Monitoring of protection effects on commercial fish species was lacking in this MPA.This study aimed at increasing local scientific knowledge by providing a baseline study about the most commercially fished families(Haemulidae,Lutjanidae and Serranidae)around Punta Francés MPA.Data collected represent only a limited period but can be used as a base point to support future monitoring.Fish abundance in number and biomass,and size were collected to test for differences between inside and adjacent areas outside the Punta Francés MPA,in different coral reef types.The main result of this study was the significantly larger size and biomass of snappers and groupers outside the MPA where intense fisheries occur.An observation consistent with a large spawning aggregation was also recorded outside the MPA.Even with a limited set of data,these results suggest that,at least temporarily,the most targeted species[and sizes]are still highly vulnerable to fisheries.Relevant habitats in the adjacent area,that are apparently missing within the MPA,may support some of the results found.Extending the limits of the Punta Francés MPA to include those habitats outside may be critical to ensure the effectiveness of this MPA in contributing to protect the most commercial species of the region.This should be done together with efficient fisheries management measures in the region,such as the significantly increase of minimum legal sizes and temporal closures during spawning migrations.

1.Introduction

Marine protected areas(MPAs)are increasingly being used as conservation and fisheries management tools.Well-performing marine protected areas(MPAs)aiming at protecting commercial or vulnerable fish species may result in considerably larger fish sizes,biomass and density within their borders compared to outside(Claudet et al.,2010;Lester et al.,2009;Zupan et al.,2018).Yet,well-performing MPAs need to be well-planned,-designed,-regulated,-managed,-enforced and-monitored MPAs to result in significant protection effects including fisheries benefits within and/or surrounding MPAs(Ahmadia et al.,2015;Gill et al.,2017;Zupan et al.,2018).

Commercial species are usually relatively mobile and may use different habitats(Claudet et al.,2010;Dorenbosch et al.,2007;Farmer et al.,2011).Hence MPAs protecting them need to be large enough to include species’home ranges and ontogenetic migrations or to be connected to other MPAs in effective networks in order to be successful(Botsford et al.,2009;Herbig et al.,2019;Russ,2002).Incorporating multiple essential fish habitats,protecting reproductive biomass and spawning aggregations of commercially targeted species may,in fact,disproportionately contribute to conservation and sustainable fisheries goals and thus,should be considered when planning and designing an MPA(Erisman et al.,2017;Farmer et al.,2011;Russel,Luckhurst,&Lindeman,2012).Scientific studies supporting MPA placement and design as well as long-term monitoring are globally underconducted but they are necessary to understand if each MPA is effective at meeting designation goals and resulting in socioecological benefits,or if changes in protection measures are needed(Osenberg,Shima,Miller,&Stier,2011).

Cuba,the largest archipelago in the Caribbean(110,860.6 km2),hosts a recognised high and unique marine biodiversity.The country has established important management measures to protect its marine and terrestrial biota and it is leading the designation of marine protected areas(MPAs)in the Caribbean region,already covering approximately 25% ofitsinsularshelf(Angulo-Valdés,Martinez,Castaneda,Frazer,&Adams,2017;Valderrama et al.,2018).Despite these measures,and similarly to many other regions in the world,most Cuban MPAs are not efficiently studied,monitored,enforced or financed(Angulo-Valdés et al.,2017;Valderrama et al.,2018).

The Punta Francés National Park,located off Isla de La Juventud,southwest of Cuba,is one of the first Cuban MPAs and a flagship of Cuban conservation programmes.Some studies were conducted to assess Punta Francés MPA(PFMPA)marine biodiversity(De la Guardia,González-Díaz,&Castellanos,2004a;b)and PFMPA relevance for recreational diving activities through the study of its most important features for diving,diving impacts on coral reefs and diving carrying capacity(Angulo-Valdés,2005,p.284;Angulo-Valdés,Acevedo,Hernández,&Sansón,2007).Yet,these studies were conducted only within the MPA,precluding comparisons between inside and outside areas.Further,an unpublished study assessing the ecological effectiveness of this MPA was performed in 1998,but the outside area considered for comparison was relatively distant and with different features(Polunin,2000).Hence,the assessment of protection effects on commercial fish species was still largely absent for the PFMPA.The present study aimed at increasing local scientific knowledge by providing a baseline study about the most fisheries-targeted fish families(Haemulidae,Lutjanidae and Serranidae)inside and outside the Punta Francés MPA.Even if this work was not included in an established monitoring programme for MPAs(non-existent),and data collected represent only a limited period,it provides information that can be used as a baseline,to support such monitoring in the future.

2.Material and methods

2.1.Study area

Punta Francés MPA(PFMPA)was proposed in 1984 due to its semipristine conditions,no permanent inhabitants within its borders,large diversity of marine species and reef types and high relevance for tourism,namely diving activities,as well as for fisheries(De la Guardia et al.,2004a).Later it was classified as a National Park(CNAP,2004,p.112)and integrated into a network of Cuban MPAs(Valderrama et al.,2018).The protection of commercial species within MPAs with expectations of biomass export to adjacent areas(spillover)to sustain local fisheries is among the goals of Cuban MPAs(Valderrama et al.,2018).

Punta Francés was used only by fishers until 1976,when it turned into a nature-based tourism region,strongly focused on diving.Fisheries and tourism were in conflict the following years and in 1996 fisheries became prohibited inside the Punta Francés MPA.Nevertheless,this is not a truly no-take area as lobster can be fished with traps(except during closure months)and small bait fish(Engraulidae)used in tuna fisheries can also be fished with small gillnets near shore(Angulo-Valdés,2005,p.284).

Even with its relatively small size(3000ha of marine area),this MPA encompasses multiple habitats,such as sandy and rocky areas,coral reefs(inner,patches,fore reef),mangroves and seagrasses.An important part of the MPA is a protected bay with healthy corals surrounded by a long sandy beach and mangroves.The north-western area,next to Francés Cape(Punta)is a more exposed area,where the insular shelf drops offinto the deeper Caribbean Sea(Fig.1;De la Guardia,2004a).

The quality of Punta Francés coral reefs sustains a large marine biodiversity including fish species and justified its designation as an MPA(De la Guardia et al.,2004a,b).Previous studies classified the coral types(biotopes)of Punta Francés MPA into two main types,wall reef(W)and spur and groove(SG).The wall reef is a gentle and relatively shallow drop-off,parallel to the coast,representing an old coastline that is now submerged.The upper part is between 8 and 10m and the bottom is at 14–15m deep(De la Guardia,Angulo,González-Sansón,Aguilar,&González-Díaz,2004b).Spur and groove biotope reaches the end of the insular shelf,which usually occurs at depths around 20m deep.This biotope is composed of longitudinal structures perpendicular to coast and spaced by sandy channels approximately 20–30m long,10–15m wide and 1–2m tall(De la Guardia et al.,2004b).Differences in fish species composition and density between these two biotopes were suggested in previous studies(Angulo-Valdés et al.,2007;González-Sansón,Aguilar,Angulo-Valdés,&González,1997;Polunin,2000),justifying the study of both biotopes to assess if potential differences between inside and outside the MPA are consistent across biotopes.

2.2.Sampling

Two fifteen-day sampling trips were conducted in April and May 2006 in the PFMPA on the R/V ‘Felipe Poey’for the Centro de Investigaciones Marinas(CIM)of Havana University.Assessing differences in biological indicators of commercial fish species between inside and outside the MPA was one among other objectives of those trips.

Data was collected at 12 different sites,8 inside the MPA and 4 outside,in the north-western adjacent area.Sampling sites were evenly distributed across the two main biotopes,wall(W)and spur and groove(SG).The locations of the sampling sites were selected randomly,according to protection level and biotopes,but were adjusted to the nearest known safe-anchoring sites/mooring buoys when existing nearby(Fig.2).

Three replicates,separated by～100m from each other,were sampled in each site and 10 individual observations(censuses)were made for each replicate,equalling 30 censuses in each site and 360 in total.Of these,342 were considered as successful and used for further analysis.

Underwater visual census(UVC)techniques are non-extractive and thus appropriate for monitoring MPAs and particularly no-take areas(Harmelin,Bachet,&Garcia,1995).In the Caribbean tropical region,fixed point counts have been shown to be more accurate than band transects(Colvocoresses&Acosta,2007).Hence,here the UVC sampling followed an adaptation of the Underwater Stationary Visual Census Technique for Quantitatively Assessing Community Structure of Coral Reef Fishes(Bonhsack&Bannerot,1986).Given that this study focused only on commercial fish species,we only recorded individuals from the Haemulidae,Lutjanidae and Serranidae families.These are considered the most important demersal commercial fish species in this region(Angulo-Valdés,2005,p.284;Baisre,2004,p.372;Claro,Sadovy de Mitchenson,Lindeman,&García-Cágide,2009).Further,a radius of 5m was used,instead of the 7.5m suggested by the original authors,to ensure similar visibility conditions at all sites.Divers trained in advance to ensure accurate assessment of both the 5m cylinder radius and fish size measurements.

In each dive,commercial fish species were identified,counted,and their total length(cm)were visually estimated,following the Bonhsack and Bannerot(1986)technique.Depth(m),water visibility(in meters),currents(simple index varying from 0–none to 5–very strong,)and bottom complexity(simple index varying from 1–low to 5–very high)were recorded.Lengths were converted to weight using weight-length(W-L)relationships from the literature(Claro&García-Arteaga,2001;Froese&Pauly,2007)and species biomass was calculated for each census.

Fig.1.Location of Punta Francés MPA,in the Youth Island(Isla de La Juventud),Cuba(map source:Google Earth 2019).

2.3.Data analysis

Ecological differences between inside and outside Punta Francés MPA and between biotopes were tested for each fish family independently. Haemulidae (grunts), Lutjanidae (snappers) and Serranidae(groupers)have distinctive ecological and morphological features and different commercial values among families,with similar life history characteristics within families(Appeldoorn&Lindeman,1985;Claro&García-Arteaga,2001;García-Cagide,Claro,&Koshelev,2001;Valdés-Mu?oz&Mochek,2001)justifying separated analyses.Also,grouping species by family ensured powerful statistical analyses.The response variables used were abundance in number(counts)and biomass(kg)and average fish length(cm).

Fig.2.Sampling sites in the study area.ONNS:Outside North North Spur and Groove;ONNW:Outside North North Wall;ONS:Outside North Spur and Groove;ONW:Outside North Wall;INS:Inside North Spur and Groove;INW:Inside North Wall;ICS:Inside Centre Spur and Groove;ICW:Inside Centre Wall;ICSW:Inside Centre South Wall;ICSS:Inside Centre South Spur and Groove;ISS:Inside South Spur and Groove;ISW:Inside South Wall(map source:Google Earth 2019).

Statistical analyses were conducted to assess potential differences in the ecological response variables between protection levels(insidevs.outside the MPA),biotopes(wallvs.spur and groove)and with the interaction between these two factors.Response variables were tested with all fish sizes together and separately for fish above and below length at maturity.Species length at maturity(Appendix A,Table A.1)were obtained from literature(Alvarez-Lajonchère,2014;Froese&Pauly,2019;García-Cagide et al.,2001).

Tests were conducted for each fish family by fitting generalized linear models(Crawley,2002,p.761;GLM,McCullagh&Nelder,1989,p.511)and corresponding ANOVAs to the data collected.These statistical tests perform well with unbalanced designs.

The family of the GLM and data transformation were chosen to ensure the best possible fit to the distributional assumptions behind the significance tests(Crawley,2002,p.761).Counts,biomass and mean length were log-transformed and GLM distribution families were Quasipoisson(counts),Gaussian(biomass)and Gamma(average length).In all cases,the null hypothesis tested was of no difference between protection levels or biotope types.Boxplots were used to visualise patterns obtained.For simplicity only plots of Biomass(kg)against protection level and biotope,with all sizes together,are displayed.

To understand differences between protection levels in the environmental variables recorded at each site(visibility,currents index and complexity index,both assumed as numeric variables),two sample Mann-Whitney non-parametric tests were run,with a two-sided alternative hypothesis.

All statistical analysis was conducted in R version 3.6.1 (R Development Core Team,2019)and RStudio version 1.2.1335.

3.Results

A total of 16070 fish from 19 species were recorded-6 grunt species(Haemulidae),7 snapper species(Lujtanidae)and 6 grouper species(Serranidae),in 342 successful individual censuses(for a list of the species recorded see Table A.1,Appendix A).A total of 19 species were recorded inside the MPA and 12 outside.A summary of the overall data collected is in Table A.2(Appendix A).

3.1.Abundance and size by fish family

3.1.1.Haemulidae

3.1.1.1.Number.Statisticalanalysisofthe observationsforthe Haemulidae species(grunts),show a significantly higher abundance in number in the biotope wall(p＜0.0001).However,there is a significant interaction between the biotope and the protection level(p＜ 0.05),suggesting significant differences in the number of grunts between inside and outside only in the spur and groove,with higher values outside the MPA(and similar values in the biotope wall).Grunts above or below length at maturity show the same patterns;they are significantly more abundant in the biotope wall(p＜ 0.01 orp＜0.0001,respectively)with no significant interaction between factors.

3.1.1.2.Biomass.The same results were found for biomass(higher in wall biotope:p＜0.001 for all sizes and below length at maturity andp＜0.01 for individuals above length at maturity;Fig.3).

3.1.1.3.Size.Values of average length are not significantly different between protection levels and biotopes,for all sizes and also for fish above or below length at maturity.

3.1.2.Lutjanidae

During exploratory analysis of the Lutjanidae data,an aggregation of large(average 75.4 cm±0.6 SE)Lujanus cyanopterus,Cubera snapper,was detected outside the protected area.A total of 405 individuals(average biomass of 343 kg±157.8 SE)was registered in one sampling site(across the three replicates=30 individual censuses)in a wall biotope(site:ONNW),close to the shelf drop-off,at sunset(18:00 local time)of May.This was considered consistent with a fish spawning aggregation.Besides this observation in this site,the total number recorded for this species was 21 individuals across six sites.This large group was considered an outlier and it was removed from the following statistical analysis assessing differences between inside and outside the MPA.

3.1.2.1.Number.The abundance in number of the Lutjanidae family(snappers)is significantly higher in the biotope wall(p＜0.05).For the individuals above length at maturity,the abundance in number is no longer significantly different between distinct biotopes.For individuals below length at maturity,significantly higher numbers are found in the wall biotope(p＜0.05).

3.1.2.2.Biomass.The biomass of snappers is highly significantly greater outside the MPA(p＜0.0001),with no interacting factors(Fig.4).The same pattern is obtained for the individuals above or below length at maturity separately;the biomass is significantly greater outside the MPA(p＜0.01 andp＜0.0001,respectively;with no interacting factors).

3.1.2.3.Size.Forthe Lutjanidae species,the average length is significantly larger outside the MPA(p＜0.05),and additionally,it is significantly greater in the biotope spur and groove(p＜0.05),with no interacting factors.For fish above length at maturity,differences are no longer significant,despite higher median lengths outside.For individuals smaller than length at maturity,significant differences are only found for the biotope,with larger sizes in the spur and groove(p＜0.01).

3.1.3.Serranidae

3.1.3.1.Number.Statistical analysis reveals that numbers of the Serranidae(groupers)are not significantly different between distinct protection levels or biotopes.Since all individuals above length at maturity were observed outside the MPA,the analysis was not carried out.For individuals below length at maturity,the same pattern was obtained(no significant differences),with the same result for all fish pooled together.

3.1.3.2.Biomass.Biomass of groupers was significantly greater outside the MPA than inside(p＜0.001;Fig.5).The same pattern is observed for individuals below length of maturity(p＜0.05).Individuals above length of maturity were only found outside the MPA.

3.1.3.3.Size.Average length of Serranidae species is highly significantly greater outside the MPA(p＜0.001).The same pattern,with a higher p-value,is obtained for individuals below length at maturity(p＜0.01).Individuals above length of maturity were only found outside the MPA.

3.2.Environmental features

When tested for differences in currents,visibility and complexity between inside and outside:a)currents were significantly different between the two protection levels(W=303532,p-value＜0.001),with higher mean values inside(due to few higher values),but the median was the same between areas.The upper quartile was,however,higher outside.b)visibility and complexity(index)were significantly higher(both median and mean)inside the MPA(W=43094 and W=103040,respectively,both withp-value＜0.001).

Fig.3.Box plots of Haemulidae log(Biomass,kg)in relation to protection level(left)and biotope(right).Box plots represent median(horizontal line),upper and lower quartiles(box limits),minimum and maximum values of the data(wiskers)and outliers(individual points).

4.Discussion

4.1.Lack of positive MPA results in abundance and size

This study suggests a lack of positive outcomes in commercial fish species abundance and size within the MPA compared to outside.Contrary to grunts(Haemulidae),snappers(Lutjanidae)and groupers(Serranidae),showed significant differences in biomass and size by protection level,but values were never higher inside.Instead,these species showed larger values in the area outside but adjacent to the Punta Francés MPA.For these species,biomass patterns are likely mostly driven by the large sizes recorded(and corresponding weights),as numbers were not significantly higher outside.Groupers and snappers are much more valuable than grunts in that region,and thus much more targeted by and vulnerable to commercial fisheries(Claro et al.,2009).That suggests most fished species[and sizes]are still highly vulnerable to fisheries outside the MPA.

An observation consistent with a fish spawning aggregation(FSA)of Cubera snapper was also located outside the MPA.Being largely predictable in time and space,FSA are highly vulnerable to fisheries,which may ultimately imperil population persistence(Erisman et al.,2012;Heyman,Kjerfve,Graham,Rhodes,&Garbutt,2005;Sadovy et al.,2008).That is why incorporating spawning aggregations.of commercially targeted species within MPAs may contribute significantly to conservation and sustainable fisheries goals(Erisman et al.,2017;Russel et al.,2012).

Hence,even with a limited set of data,not representing the temporal dynamics of the species assessed,these results suggest that,at least temporarily,the most targeted species are not benefiting from adequate MPA protection.

Fig.4.Box plots of Lutjanidae log(Biomass,kg)in relation to protection level(left)and biotope(right).＊Not including the aggregation of Cubera snapper.Box plots represent median(horizontal line),upper and lower quartiles(box limits),minimum and maximum values of the data(wiskers)and outliers(individual points).

4.2.Habitat features

Potential habitat confounding effects may explain some of the outcomes obtained.Polunin(2000),in a previous study assessing protection effects in Punta Francés found distinct and,in some cases,contradictory results than those of the present study.Higher numbers and biomass of groupers and grunts,and higher numbers of snappers were found inside the MPA compared to outside(Polunin,2000).Yet,the outside area was in the opposite direction(south-eastern tip of Youth Island,Punta del Este)of the one studied here(adjacent to the northwestern border),precluding direct comparisons.Polunin(2000)suggested a possible habitat confounding effect on protection responses,influenced by the large distance between the MPA and the outside area assessed.Despite the proximity between the two areas in the present study,habitat variations(at different scales)may also play a role in distinct species distribution,resulting in the observed higher biomasses of snappers and groupers outside.Density of groupers was shown to be correlated to physical and biotic habitat attributes,additionally to reef biotope(Sluka,Chiappone,&Sullivan Sealey,2001).Here,habitat relevance is suggested by biotope influence in some indicators and species(grunts and snappers),as well as by the significant interaction found between biotope and protection level in some cases.Yet,other environmental features besides bottom-type biotope are likely contributing to the patterns found.

The majority of the MPA is surrounded by a quiet bay,whereas the area next to Frances Cape(north-western MPA border)is in the insular shelf drop-off,with potential for(and observed)stronger currents,lower visibility and more exposed conditions.In fact,according to De la Guardia et al.(2004b)[Fig.2,where bottom types are mapped],some habitat differences may occur between the lagoon and the area next to the MPA border.

Fig.5.Box plots of Serranidae log(Biomass,kg)in relation to protection level(left)and biotope(right).Box plots represent median(horizontal line),upper and lower quartiles(box limits),minimum and maximum values of the data(wiskers)and outliers(individual points).

Punta Francés MPA features are key to support the primary goal of this MPA establishment which is biodiversity conservation for tourism purposes(e.g.recreational diving)(De la Guardia et al.,2004a,b;Angulo-Valdés et al.,2007).By including different adjacent habitat types,this MPA may contribute to the protection of grunts and snappers moving between corals(day)and mangroves and seagrass meadows nearby(night)to feed on invertebrates(Valdés-Mu?oz&Mochek,2001).Meyer,Schultz,and Helfman(1983)suggested that this nocturnal migration is an important source of energy exchange among adjacent habitats and may increase coral growth rate considerable.In tropical regions,reef fish commonly use and depend on multiple habitats,moving between coral reefs,seagrasses and mangroves,with some species displaying ontogenetic habitat shifts(Aguilar,González-Sansón,Cabrera,Ruiz,&Curry,2014;Honda,Nakamura,Nakaoka,Uy,&Fortes,2013).In fact,in coral reefs,abundance and size patterns of some fish species are probably related to the proximity of mangroves and seagrasses(Dorenbosch,Verberk,Nagelkerken,&van der Velde,2007;González-Sansón,Aguilar,Hernández,Cabrera,&Curry,2008).These habitats are recognised as critical nurseries for coral reef fish species(Terrados&Borum,2000)and have unique functional values in the ecosystem(Harborne et al.,2006).Thus,daily or ontogenetic connectivity between these habitats,as well as exclusive usage of a single habitat,justify the inclusion of multiple habitats within MPAs(Honda et al.,2013),as in the PFMPA.

However,due to its location,Punta Francés MPA is apparently missing important habitats(less attractive to recreational divers)that are found in the assessed areas adjacent to the MPA.Similar reef types(considered here as biotopes)may be associated with distinct conditions resulting in different seascapes.These may be used differently by the various species surveyed,possibly explaining the results found.

Other MPA studies in the Caribbean and other tropical regions,report that groupers may easily benefit from protection due to their typically territorial behaviour with limited home ranges(Chiappone,Sluka,&Sealey,2000;Eklund,2000;Unsworth,Powell,Hukom,&Smith,2007).Even snappers,showed to benefit from protection in a few Caribbean MPAs(Polunin&Roberts,1993).Hence,even a relatively small MPA as PFMPA could potentially lead to protection effects if including some missing essential habitats as those occurring adjacently to the PFMPA.This could be addressed by extending the MPA area towards the north-west.

4.3.Fisheries

The main result of this study points to the vulnerability to fisheries of the most valuable fish families and sizes recorded outside this MPA.The low abundance of large snappers and groupers within the PFMPA was already reported by Angulo-Valdés et al.(2007).They did not survey adjacent areas outside,precluding the comparison between protection levels.Yet,they justified their findings with the illegal fishing within the MPA and with relatively intense commercial fisheries outside,adjacent to the MPA borders(also known as ‘ fishing the line’;Angulo-Valdés et al.,2007).

In fact,the outside area of the present study is in the Gulf of Batabanó,southwest region of Cuba,the largest Cuban fishing area(Claro et al.,2009;Claro&Lindeman,2003;De la Guardia et al.,2018).The western region of Gulf of Batabanó is known and used by fishers tracking migration routes and spawning aggregation sites,with respective species showing over fishing and an urgent need for effective fisheries management measures(Claro&Lindeman,2003;De la Guardia et al.,2018;Valle&Ramos,2013).

The surveyed area is subjected to commercial fisheries using lines traps,gillnets,small mid-water trawl nets and small artisanal fixed traps “almadravas”,where one end of the entrance is attached to the mangrove roots and the other extends in the opposite direction(personal observation;Angulo-Valdés,2005,p.284).These gears can be deployed at different depths and habitats,with gillnets and traps being the main commercial fishing gears at the depths and habitats surveyed.Fishing gears are deployed to intercept species movements between habitats(mangrove,seagrass meadows,coral reefs)or between the insular shelf and the drop-off,movements that increase during reproductive migrations(Baisre,2004,p.372;Claro et al.,2009;Claro&Lindeman,2003).

Despite the high potential of fishing pressure outside the MPA,biomass and size of groupers and snappers were larger outside,rather than inside.Spillover could occur,but it is only assumed if related to higher numbers and biomass inside(Abesamis,Russ,&Alcala,2006;Gell&Roberts,2003).These individuals may move between different habitats or depths(Claro&Parenti,2001;Farmer et al.,2011;Goldstein,D’Alessandro,&Sponaugle,2016;Herbig et al.,2019;Tupper&Rudd,2002;Valdés-Mu?oz&Mochek,2001),and possibly momentarily benefit from protection within the MPA with opportunity to grow.Yet,results highlight that,at least temporarily,the most targeted species[and sizes]are still highly vulnerable to fisheries.Groupers above length at maturity were not found within the MPA.Further,even groupers and snappers below length at maturity,showed significantly larger sizes and biomasses outside,suggesting that the role of this MPA in protecting juveniles has yet to be demonstrated.

Some of the grouper and snapper species observed are reported as critically endangered(Epinephelus striatus),vulnerable(Lutjanus cyanopterus)or near threatened(Mycteroperca bonaciandM.venenosaorL.analisandL.synagris)by IUCN(2019),all with a decreasing pattern.Further,a recent paper studying the over fishing of snapper populations in the region of Batabanó(De la Guardia et al.,2018),reported that the minimum legal size ofL.analisandL.cyanopterusare half of their lenght at maturity,which combined with fishing of FSA,lead to growthand recruitment-over fishing,preclude spawning and population persistence.

4.4.Fish spawning aggregations

The region of Punta Francés,and specifically the surveyed area outside,where the insular shelf drops offinto the deeper Caribbean Sea,is locally described as an area where fish species migrate to during spawning aggregations(Angulo-Valdés,2005,p.284;local fishersin com.pess.,2006).Ecologically,reef drop-offs are hotspots of reproduction for tropical fish species(Coleman,Scanlon,&Koenig,2011;Farmer et al.,2017;Koening et al.,2000).Indeed,in Cuba,most locations described as spawning aggregation sites are near the shelf dropoff at 20–50m(Claro&Lindeman,2003).

Snapper peak spawning times are known to occur between April and August in this region(De la Guardia et al.,2018;García-Cagide et al.,2001),which is in accordance with our observation of a vast group of Cubera snappers.Other spawning aggregations of this species in Cuba were described to occur at night between June and September at around 20–30m(Claro&Lindeman,2003),which are relatively similar conditions to what we have observed.

Another MPA located in the same region of Gulf of Batabanó but further west,surrounding the Cayo San Felipe(San Felipe Keys National Park,SFKNP)was established in 2010,after the discovery of fish aggregation sites ofL.analis(Mutton snapper)andL.synagris(Lane snapper)in 2001(Claro&Lindeman,2003),but the MPA did not include the area where these aggregations were observed(De la Guardia et al.,2018).Similarly,an aggregation of Nassau grouper(Epinephelus striatus),a critically endangered species,was described around 3 km west of Punta Francés MPA border,not being included within the MPA(Claro&Lindeman,2003).These patterns suggest spawning sites are being left out of some Cuban MPAs,despite the recognition that Cuban networks of MPAs are to be established incorporating information on spawning sites(Claro&Lindeman,2003;CNAP,2013).

To the best of our knowledge,no previous studies described any spawning aggregation of Cubera snapper adjacent to the Punta Francés MPA.Direct observations of spawning aggregations are very rare and difficult to obtain,justifying the indirect method(e.g.fishers'knowledge)used by Claro and Lindeman(2003)to identify grouper and snapper spawning sites in Cuba.Such characteristics also justify the report of this[rare]observation.Its features are consistent with a spawning aggregation and merits attention.This finding is particularly relevant because i)no Cubera snapper spawning sites were reported in the intensively fished region of Gulf of Batabanó(Claro&Lindeman,2003);ii)spawning sites are known to be predictable in time and space by fishers(Erisman et al.,2012;Farmer et al.,2017;Heyman et al.,2005;Russel et al.,2012),and thus one observation may reveal‘the vulnerable site’;iii)protecting spawning biomass and respective spawning sites,as the one recorded for Cubera snapper,may disproportionally contribute to positive MPA effects(Erisman et al.,2017);iv)Cubera snapper is considered to be over fished in Cuba,due to growth-and recruitment-over fishing and FSA-directed fisheries(Claro et al.,2009;De la Guardia et al.,2018);and v)it was assessed as Vulnerable by the IUCN red list,with a decreasing pattern(IUCN,2019).

5.Conclusions

Although one of the main objectives of Cuban MPAs is to protect commercial fish species,this is one of the few MPA-related studies on the potential protection effects.It is therefore a relevant contribution that may support current efforts in Cuban marine conservation by providing a baseline study that can be compared and support monitoring in Punta Francés MPA.

This MPA might be accomplishing its main goal of protecting biodiversity for tourism purposes.Yet,the fact that adjacent to this MPA,heavily fished species[and sizes]are still highly vulnerable to fisheries and apparently not effectively benefiting from protection measures inside,calls for attention and should be addressed.This is largely reinforced by the observation consistent with a spawning aggregation outside the MPA.

Some of the snapper and grouper species surveyed may benefit from relatively small MPAs if well placed.Punta Francés MPA could potentially lead to consistent protection effects if including relevant habitats such as those occurring adjacent to the MPA.Also,including spawning sites,as the one recorded for Cubera snapper,would likely contribute to positive protection effects of this MPA.Ecological differences between inside and outside the MPA are possibly largely explained by different seascapes,with an exposed insular shelf drop-off characterizing the outside and a quiet bay the inside area.By extending the MPA limits towards the north-west,a potential essential fish habitat and a recorded spawning site could be included within the MPA.This study suggests such rezoning is necessary as well as the development of a monitoring program for MPAs.

Global and regional patterns previously described for the studied species highlight the urgent need for efficient management and protection measures.Even if MPAs alone are not enough to protect these species,they may provide significant contributions if properly placed[and enforced]while combined with widespread adequate fishery management measures and spatial planning.

Additional measures should be adopted in Cuban waters for the most vulnerable snappers and groupers,such as the substantial increase of the minimum legal size and the establishment of temporal closures in the Golf of Batabanó during spawning migrations.

Declaration of competing interest

None.

Acknowledgements

We thank CIM(University of Havana)and University of Algarve.Special thanks to ‘Felipe Poey’crew,who become special friends:Orestes,Aramis and Yeyo and to Nene:diver and nature vigilant of PFMPA.Thanks to some special local fishermen who helped a lot understanding local fisheries and become friends.Thanks to all technicians and friends who made this work possible.B.H.C.was supported by Portuguese national funds through FCT-Foundation for Science and Technology,I.P.,in the scope of Norma Transitória DL57/2016/CP1361/CT0038.This study also received Portuguese national funds from FCT through Project UID/Multi/04326/2019.

Appendix A

Table A.1List of the commercial species(and respective family)recorded in this study through UVC within and outside Punta Francés MPA.Maximum size of the species(cm;obtained from FishBase.org,Froese&Pauly,2019),length at first maturity(Lmat,cm;García-Cagide et al.,2001;Alvarez-Lajonchère,2014;Froese&Pauly,2019)and IUCN red list category(including trend)are also shown.LC:least concern,DD:data deficient,NT:near threatened,VU:vulnerable,EN:endangered,CR:critically endangered,EW:extinct in the wild,EX:extinct(IUCN,2019).

Table A.2Main response variables obtained for each commercial species recorded by UVC in Punta Francés MPA and adjacent area,by protection level and bottom type:Average number(counts),Average length(cm),Average biomass(kg)and respective Standard Errors(±SE).Out:outside,In:Inside,SG:spur and groove,W:wall.Counts and biomass were summed within each site and site values were averaged across protection level and biotope.Standard errors=NA means only recorded in one site.

Protection Level Biotope Family Species Av.Number Av.Length_cm Av.Biomass_kg Inside SG Lutjanidae Lutjanus mahogoni 143.3(±118.3) 15.0(±0.7) 10.7(±9.5)Inside SG Lutjanidae Lutjanus synagris 1(±0) 16.5(±1.5) 0.08(±0.02)Inside SG Lutjanidae Ocyurus chrysurus 203.8(±56.5) 14.3(±0.5) 23.3(±6.6)Inside SG Serranidae Epinephelus adscensionis 3(±2.0) 19.5(±0.5) 0.3(±0.1)Inside SG Serranidae Epinephelus guttatus 3(±2.0) 10.7(±0.7) 0.05(±0.03)Inside SG Serranidae Epinephelus striatus 5.3(±1.4) 20.6(±1.8) 0.7(±0.2)Inside SG Serranidae Mycteroperca bonaci 1.3(±0.3) 23.5(±2.6) 0.3(±0.2)Inside SG Serranidae Mycteroperca tigris 2(±1.0) 21.8(±4.4) 0.4(±0.2)Inside SG Serranidae Mycteroperca venenosa 2(±NA) 14.5(±NA) 0.1(±NA)Inside W Haemulidae Anisotremus virginicus 6.5(±2.5) 15.9(±0.9) 5.8 ± 0.5)Inside W Haemulidae Haemulon album 3(1.0) 16(±1) 3.1(±2.2)Inside W Haemulidae Haemulon flavolineatum 1080(±403.4) 10.9(±0.1) 29.7(±13.6)Inside W Haemulidae Haemulon parra 3(±NA) 20.3(±NA) 0.4(±NA)Inside W Haemulidae Haemulon plumieri 687.8(±305.3) 13.9(±0.7) 35.1(±17.0)Inside W Haemulidae Haemulon sciurus 478.8(±254.6) 12.9(±0.2) 21.7(±11.4)Inside W Lutjanidae Lutjanus analis 1(±NA) 70(±NA) 6.3(±NA)Inside W Lutjanidae Lutjanus apodus 56.8(±15.0) 12.7(±0.5) 2.2(±0.7)Inside W Lutjanidae Lutjanus cyanopterus 2(±NA) 80(±NA) 5.8(±NA)Inside W Lutjanidae Lutjanus mahogoni 21(±9.0) 11.0(±0.3) 0.5(±0.2)Inside W Lutjanidae Lutjanus synagris 7(±NA) 9.9(±NA) 0.1(±NA)Inside W Lutjanidae Ocyurus chrysurus 224(±36.3) 12.9(±0.2) 18.7(±4.1)Inside W Serranidae Epinephelus guttatus 1(±NA) 17(±NA) 0.1(±NA)Inside W Serranidae Epinephelus striatus 2.3(±0.8) 15.6(±0.6) 0.1(±0.03)Inside W Serranidae Mycteroperca bonaci 2.8(±0.6) 25.1(±5.5) 1.7(±1.0)Inside W Serranidae Mycteroperca tigris 1(±NA) 14(±NA) 0.04(±NA)Inside W Serranidae Mycteroperca venenosa 2.7(±1.7) 15(±0.6) 0.1(±0.1)Outside SG Haemulidae Haemulon flavolineatum 2(±NA) 13.5(±NA) 0.1(±NA)Outside SG Haemulidae Haemulon plumieri 223.5(±94.5) 14.7(±0.1) 11.6(±3.9)Outside SG Lutjanidae Lutjanus analis 5(±NA) 68(±NA) 16.5(±NA)Outside SG Lutjanidae Lutjanus cyanopterus 4(±2.0) 56.9(±7.8) 5.5(±1.0)Outside SG Lutjanidae Lutjanus jocu 1(±NA) 55(±NA) 3.1(±NA)Outside SG Lutjanidae Ocyurus chrysurus 188(±NA) 12.8(±0) 13.4(±NA)Outside SG Serranidae Epinephelus adscensionis 4.5(±0.5) 26.5(±1.8) 1.2(±0.3)Outside SG Serranidae Epinephelus guttatus 7.5(±4.5) 16.9(±0.8) 0.4(±0.3)Outside SG Serranidae Epinephelus striatus 6(±1) 33.9(±4.3) 10.6(±5.9)Outside SG Serranidae Mycteroperca bonaci 4.5(±2.5) 59.6(±16.1) 17.6(±16.3)Outside W Haemulidae Haemulon flavolineatum 1207(±NA) 11.6(±NA) 40.6(±NA)Outside W Haemulidae Haemulon plumieri 233(±211) 14.7(±0.4) 13.1(±11.6)Outside W Lutjanidae Lutjanus analis 3(±NA) 35(±NA) 4.3(±NA)Outside W Lutjanidae Lutjanus apodus 2(±NA) 10(±NA) 0.03(±NA)Outside W Lutjanidae Lutjanus cyanopterus 405(±NA) 75.4(±NA) 1024.2(±NA)Outside W Lutjanidae Lutjanus jocu 30(±NA) 39.8(±NA) 45.8(±NA)Outside W Lutjanidae Ocyurus chrysurus 178.5(±57.5) 14.5(±0.8) 19.3(±1.0)Outside W Serranidae Epinephelus adscensionis 4(±1) 19.2(±4.2) 0.5(±0.2)Outside W Serranidae Epinephelus guttatus 12(±NA) 13.2(±NA) 0.6(±NA)Outside W Serranidae Epinephelus striatus 2.5(±1.5) 35(±5) 3.1(±2.9)Outside W Serranidae Mycteroperca bonaci 3(±NA) 81.7(±NA) 38.0(±NA)Outside W Serranidae Mycteroperca venenosa 2(±NA) 20(±NA) 0.3(±NA)