Coral reef fish assemblages exhibit signs of depletion in two protected areas from the eastern of Los Canarreos archipelago (Cuba, Caribbean Sea)

Introduction

The designation of marine protected areas (MPAs) is a well-known approach to counteract the global deterioration of coastal ecosystems due to synergistic stressors, mainly fisheries and eutrophication (Hughes et al., 2003; Hoegh-Guldberg & Bruno, 2010; Costanza et al., 2014; Harborne et al., 2017). Indeed, successful MPAs are connected to the recovery of biological populations (Roberts et al., 2001; Kelaher et al., 2014; Strain et al., 2019). However, most MPA’s interventions have suboptimal or poor results and a large fraction of MPAs worldwide suffer pouching and low levels of enforcement (Mora et al., 2006; Costello & Ballantine, 2015). Accordingly, MPA categories with different levels of fishing restrictions exist worldwide based on the criteria of the International Union for the Conservation of Nature (IUCN).

In Cuba, MPAs belonging to the National System of Protected Areas (SNAP) include a considerable portion of coastal ecosystems such as coral reefs (ca. 30%), seagrass beds (ca. 24%), and mangroves (ca. 35%) (Perera-Valderrama et al., 2018). These MPAs are classified after eight management categories, which have been adapted from IUCN categories (Perera-Valderrama et al., 2018). Two of these categories are relevant in our study: Ecological Reserve and Fauna Refuge. Ecological Reserve is equivalent to category II from the IUCN, and it protects important flora and fauna, as well as either complete or partial ecosystems, and can support nature-based tourism activities, such as low-impact hotels (CNAP, 2013; Perera-Valderrama et al., 2018). Fauna Refuge is equivalent to category IV from the IUCN. It conserves flora and fauna species and (or) populations, which can include management of habitats and ecosystems (CNAP, 2013; Perera-Valderrama et al., 2018). In theory, Ecological Reserve requires a stronger management (including fishing restrictions) and higher level of protection than Fauna Refuge; therefore, healthier fish assemblages should be expected in the former compared to the latter.

Not all Cuban MPAs have been equally studied, even despite some of them belong to the most restricted MPA categories (Navarro-Martínez & Angulo-Valdés, 2015). In this contribution, we focus on coral reef fish assemblages from two poorly studied MPAs in Los Canarreos archipelago (southwestern Cuba) (Fig. 1): Cayo Campos-Cayo Rosario Fauna Refuge (hereafter, CCCR) and Cayo Largo Ecological Reserve (hereafter, CL). Cayo Largo is a well-known tourist destination where recreational fisheries and SCUBA diving are carried out with success, depending partially on features of the fish assemblages. An integrative survey in Cayo Largo in 1998–1999 (Alcolado, Claro-Madruga & Martínez-Daranas, 2001) reported a significant deterioration of coral reef ecosystem. Posteriorly, marine biodiversity was studied in a coastal lagoon nearby to the Cayo Largo coral reef system (Guardia, González-Sansón & Aguilar, 2003), where human activities were likely not damaging the ecosystem.

Several broad-scale surveys in Cuban coral reefs included sites from Los Canarreos archipelago; namely: benthic communities (Alcolado et al., 2010, 2013; Caballero-Aragón et al., 2019, 2020), mesophotic biodiversity (Reed et al., 2018), and mesophotic fish assemblages (Cobián-Rojas et al., 2021). In all of these surveys, the ecological metrics in Los Canarreos’s sites indicated rather regular conditions respect to other coral reef sites. However, no scientific information focused on the ichthyofauna in the MPAs CCCR and CL have been published.

Fishes represent excellent indicators of ecosystem health and MPA effectiveness, mainly in coral reefs (Schmitter-Soto et al., 2017; Strain et al., 2019; Rojo, Anadón & García-Charton, 2021). Functional traits of fishes are related to ecological processes in coral reefs such as secondary productivity, nutrient cycling, and trophic controls (Bellwood et al., 2004, 2019; Lefcheck et al., 2019). Concordantly, reef fishes are likely one of the best studied assemblages in the Caribbean (Miloslavich et al., 2010), but given the increasing pressure due to fisheries and habitat deterioration, more studies are urgently needed.

The present work is aimed to compare the ichthyofauna structure in three areas in the eastern part of Los Canarreos archipelago with different management schemes: CCCR, CL, and non-protected area (nMPA), and considering habitat differences and depth variation. The biological data were collected during an expedition carried out along Los Canarreos archipelago in November 2015, as part of a comprehensive study targeted on fishes, benthic and microbial communities (Weber et al., 2019).

Materials and Methods

Results

Taxonomic structure

Over the 12 reef sites differing in MPA status and habitat type, we recorded 15,059 fishes, of which 14,794 (98%) were identified mostly to species level. We reported 80 species, three genera (Calamus, Kyphosus, and Pterois), and the species complex Acanthurus bahianus/Acanthurus tractus (Table S1). For the quantitative analyses, we used the genera and the species complex as equivalent to species resulting in 84 entries. The species belong to 24 families, with Haemulidae, Pomacentridae, Labridae, Acanthuridae, Scaridae, and Lutjanidae the best represented with more than 1,000 individuals and seven species (excepting Acanthuridae). The most abundant species were (in brackets the number of individuals): Haemulon sciurus (2,230), Thalassoma bifasciatum (1,613), Acanthurus coeruleus (1,291), and Haemulon flavolineatum (1,004).

Fish abundance varied significantly between sites with a mean of 113 ind. 125 m⁻² (range: 18–717 ind. 125 m⁻²) (Fig. 2A). However, no differences were detected between MPA category or between habitats (Table 2). Species density varied significantly between sites with a mean of 16 species 125 m⁻² (range: 5–30 species 125 m⁻²). However, habitat type explained an amount of variance (19%) similar to site (21%) suggesting that it is an important source of variation as well (Table 2). Actually, species density tends to be higher in backreef compared to forereef (Fig. 2B).

Table 2:

Results of the permutational analysis of variance (PERMANOVA) on ecological variables of fish assemblages from Los Canarreos archipelago.

Mixed model with three factors and one covariate: MPA category (three levels; fixed effect) crossed with HABITAT (two levels; fixed effect), and SITE nested within HABITAT (12 levels; random effect). DEPTH was set as covariate and transects were used as replicates.

Variable	Source of variation	Degrees of freedom	Pseudo-F	p-value	Component of variation (%)
Abundance	MPA	2	1.55	0.28	11
	HABITAT	1	1.60	0.25	10
	DEPTH	1	0.18	0.68	0
	MPA × HABITAT	2	1.43	0.30	14
	SITE (HABITAT)	6	5.86	0.0001	27
	Residual	118			38
Species density	MPA	2	2.78	0.13	16
	HABITAT	1	4.42	0.07	19
	DEPTH	1	0.10	0.76	0
	MPA × HABITAT	2	1.24	0.35	9
	SITE (HABITAT)	6	4.44	0.0006	21
	Residual	118			36
Threatened species	MPA	2	6.11	0.03	14
	HABITAT	1	1.49	0.25	4
	DEPTH	1	3.50	0.09	14
	MPA × HABITAT	2	3.51	0.09	14
	SITE (HABITAT)	6	0.47	0.83	0
	Residual	118			55
Biomass	MPA	2	1.68	0.28	13
	HABITAT	1	1.38	0.29	8
	DEPTH	1	0.16	0.68	0
	MPA × HABITAT	2	0.64	0.61	0
	SITE (HABITAT)	6	2.86	0.02	24
	Residual	118			54
Functional group richness	MPA	2	1.14	0.39	6
	HABITAT	1	0.91	0.37	0
	DEPTH	1	1.88	0.19	18
	MPA × HABITAT	2	1.11	0.40	8
	SITE (HABITAT)	6	5.82	0.0001	28
	Residual	118			40
Functional redundancy	MPA	2	2.82	0.13	17
	HABITAT	1	4.54	0.08	19
	DEPTH	1	0.11	0.73	0
	MPA × HABITAT	2	1.24	0.36	9
	SITE (HABITAT)	6	4.32	0.0005	21
	Residual	118			35
Multivariate structure	MPA	2	2.55	0.0016	13
	HABITAT	1	5.84	0.0002	19
	DEPTH	1	1.10	0.37	4
	MPA × HABITAT	2	1.79	0.024	14
	SITE (HABITAT)	6	4.26	0.0001	18
	Residual	118			31

DOI: 10.7717/peerj.14229/table-2

Note:

Significant values at p < 0.05 are in bold.

We detected ten species which are classified in four threatened categories according to IUCN (2022): Epinephelus striatus as critically endangered; Aetobatus narinari as endangered; Ginglymostoma cirratum, Lachnolaimus maximus and Lutjanus cyanopterus as vulnerable; Balistes vetula, Hypanus americanus, Lutjanus analis, Mycteroperca venenosa, and Scarus guacamaia as near threatened. The abundance of threatened species was significantly different only between MPA categories (Table 2), with the least abundance located in the CL-backreef (Fig. 3A). The most abundant threatened species were Lachnolaimus maximus, Balistes vetula, and Epinephelus striatus with averaged abundance of 0.24, 0.11 and 0.08 ind. 125 m⁻², respectively. Backreefs in Cayo Largo had only one threatened species with very low abundance, conversely forereefs in nMPA had six threatened species with higher abundance (Fig. 3B).

Biomass was significantly variable between sites, and no other clear trends emerged between MPA categories or between habitat types (Fig. 4A). The families with higher biomass were Lutjanidae and Haemulidae, followed by Scaridae (Fig. 4B).

Functional structure

Twenty-seven functional groups were observed during the study (Table S1, Fig. S2). Most redundant functional groups, according to the number of species (in brackets), were the roving browsers (six), roving grazers (six), roving predators (eight), and roving macroinvertivores (15). Conversely, large roving browsers, small cryptic macroinvertivores, pelagic macroinvertivores, midwater macroinvertivores, roving sand macroinvertivores, cryptic macroinvertivores, and pelagic planktivores were scarcely represented by one species observed in just one site. The most abundant functional groups were the roving macroinvertivores followed by the roving grazers, cryptic grazers and roving planktivores (Fig. S2).

Functional group richness varied significantly between sites (Table 2), with a mean of nine functional groups (range: 4–14 functional groups) (Fig. 5A). Functional redundancy also varied significantly between sites with a mean of 0.59 species per functional group (range: 0.19–1.07 species per functional group). Site explained 21% of the total variance, similar to habitat (19%) and MPA (17%), suggesting that all of them influenced the functional redundancy (Table 2). Actually, redundancy tended to be higher in backreefs, but only within MPAs (i.e., in CL and CCCR) (Fig. 5B).

Multivariate structure

The multivariate structure varied significantly after several factors, listed according to the amount of explained variance: Habitat, site, interaction MPA × habitat, and MPA category (Table 2). The ordination by NMDS showed that habitat (and some extension depth) causes some grouping of sites (Fig. 6).

Discussion

This research contributes to the ichthyofauna knowledge from a poorly explored coral reef region of the Caribbean Sea, namely the eastern part of Los Canarreos archipelago. We provide, for first time for this region, published comprehensive information regarding the structure of coral reef fish assemblages from a taxonomic (abundance, density and biomass) and functional (richness and redundancy) perspectives, and how these features relate to MPA category, habitat type and depth. We acknowledge that our study lacks temporal replication (i.e., a single expedition) that limits the reach of the inference. Nevertheless, our study still provides new insights into the effect of MPA condition to the reef system, and the effects of ecological drivers on the fish assemblages.

We reported that MPA’s condition did not have a significant effect over the fish assemblage structure and ecological metrics likely because the poor management. Although CCCR was approved as MPA since 2012 (Consejo de Ministros, 2012), it confronts, as other Cuban MPAs, difficulties in enforcement, mainly due to limited resources and conditions (Perera-Valderrama et al., 2018). In addition, CL lacked an administration staff and enforcement.

In this regard, only the abundance of threatened fish species and the multivariate structure of fishes, showed significant differences based on MPA status. Threatened species are those whose population size and geographic range have been reduced, usually as result of habitat deterioration and overfishing (IUCN, 2012); and they are useful indicators of protection effectiveness in MPAs and habitat availability (Navarro-Martínez et al., 2022). However, despite the significant MPA effect, threatened species were scarcely represented in most of our survey sites. Paradoxically, the abundance of these species was the lowest within CL, which is in accordance with the lack of enforcement already mentioned, at the time that represents a concern; since this MPA was identified due to its natural resources compatible with the touristic activities developed there.

Two key metrics of fish assemblages frequently used to evaluate the effectiveness of MPAs are the biomass, and the occurrence of particular functional groups (e.g., large piscivores and predators). Both of these metrics were consistently similar (i.e., not significantly different) between MPA categories. The biomass of ecologically and commercially important families showed a contrasting panorama along survey sites. For instance, we found very low biomass values in some survey sites (mostly in CL; generally <2 kg 125 m⁻²). However, several multispecies schools (Lutjanidae, Haemulidae and Carangidae) were observed along other survey sites resulting in high biomass (>12 kg 125 m⁻²) in both protected and non-protected sites. The fish families Lutjanidae, Haemulidae, Carangidae, Serranidae, and Scaridae usually show scarce biomass in deteriorated coral reefs where fisheries are allowed or MPA’s enforcement is ineffective (Valdivia, Cox & Bruno, 2017, Duran et al., 2018, Navarro-Martínez et al., 2022). This fact responds to their ecological role and to their historical extraction in non-protected areas through fisheries, including Cuban waters (Baisre, 2018). Therefore, these fish families are important proxies of healthy/protected coral reefs. In this context, very low density and biomass of herbivorous (0.02–0.03 ind. m⁻²; 2.4–3.3 g m⁻², respectively) and carnivorous (includes to Lutjanidae and Serranidae: 0.002–0.004 ind. m⁻²; 0.02–0.8 g m^-2, respectively) fishes were recorded in surveys done in Cayo Largo reef crest in 2012 (Alcolado et al., 2013). Our records of biomass for these groups were higher near habitats/sites in CL and nMPA backreefs (for comparisons, we converted our values from kg 125 m⁻² to g m⁻², with the following result, Scaridae: 11.36 g m⁻², 32.41 g m⁻²; Serranidae: 0.04 g m⁻², 2.30 g m⁻², Lutjanidae: 15.24 g m⁻², 28.96 g m⁻², respectively), but direct comparisons should be avoided since both studies included different survey techniques and survey sites. On the other hand, our current results of biomass for ecologically and commercially important families are comparable but mostly lower than other records from Cuba and the Greater Caribbean (Table 3).

Several functional groups were scarcely represented along the study region; in particular, larger individuals of Serranidae and Scaridae. These families are frequently evaluated in studies of MPA effectiveness due to their sensibility to anthropogenic impact and their importance in the ecosystem functioning (Valdivia, Cox & Bruno, 2017; Lefcheck et al., 2019). Larger groupers were barely represented by only four species with low abundance: two piscivore species of Mycteroperca and two predator species of Epinephelus genera. Similarly, larger browser parrotfishes (i.e., herbivores) were barely represented in terms of abundance.

Our results suggest that the ichthyofauna from CCCR and CL MPAs exhibit signs of depletion such as low biomass and loss of key groups. This occurs despite that they have been protected from fisheries since the ZBREUP’s creation; that is, since 1997 in CL only (MIP, 1997) and since 2012 (Ministry of Food Industry of Cuba (MINAL), 2012) for all the study area (Fig. 1). In principle, coral reefs closer to the touristic facilities at Cayo Largo should suffer the highest impact due to pollution (chemical and acoustic), physical disturbance and pouching, while the remoteness of the other sites (e.g., CCCR and non-MPA 12) hampers the characterization of anthropogenic impacts and illegal activities, which in turn limits the management. For instance, poor management signals have been reported in other studies in Los Canarreos, which reported illegal fishing in Cayo Largo and their proximities (Alcolado, Claro-Madruga & Martínez-Daranas, 2001; Guardia & González-Díaz, 2002; Azanza et al., 2018). Pouching explains the lack of larger sized fishes and depletion of their abundance in most of our study sites. This result suggests the low representativeness of critical functions to the coral reef ecosystem such as herbivory and transference of energy to top trophic levels.

Importantly, synergetic factors are likely affecting different levels of trophic webs in this area of Los Canarreos. For instance, high density of macroalgae coverage was reported as early as 1998–1999 by Alcolado, Claro-Madruga & Martínez-Daranas (2001) likely because of the low density of herbivores. Recently, a comparison between coral reefs from Florida Keys (USA), Jardines de la Reina (Cuba) as well as the same sites we surveyed in this study, reported comparably high algae and low coral coverage along Los Canarreos (Weber et al., 2019). Similarly, Caballero-Aragón et al. (2019) found rather medium-low estimates of coral metrics (diversity and living coral cover) and high mortality in Los Canarreos, compared with other Cuban reefs. Unfortunately, these results fit into the general trend of coral reef deterioration worldwide (Hughes et al., 2017).

On the other hand, fish assemblages of the region possess some favorable features typical of healthy habitats. For instance, assemblages were taxonomically and functionally diverse with 84 species and 27 functional groups, indicating high diversity and complementarity of functions that are associated with healthy coral reefs (often within successful MPAs) (Bellwood et al., 2004, 2019; Micheli et al., 2014). In addition, a few sites showed higher biomass of important families (e.g., Lutjanidae, Carangidae, Haemulidae) than others from the Greater Caribbean (Vallès & Oxenford, 2014) and similar biomass as in the successful Jardines de la Reina National Park (Navarro-Martínez et al., 2022) (Table 3). Nevertheless, considering the rather deficient management history from these areas (see above), such “high” diversity features are likely remnants of the coral reef fish assemblages that used to inhabit the region (Jackson, 1997).

In this context, the current habitat heterogeneity is playing a critical role preserving and structuring the reef fish assemblages. While MPA category and depth lacked significant effects for most of the tested metrics, our results suggest an important effect of sites (nested within habitats) on the fish assemblage structure. The significant effect of sites points to the important role of local environmental features on ichthyofauna, which are often elusive to identify (e.g., bottom rugosity, enhanced primary production). Therefore, in addition to our quantitative data about fish assemblage structure, we analyzed the underwater footage for reef characteristics and the geomorphology of the archipelago to determine the influence of cays or other habitats. We think that at least two ecological drivers may be important in shaping the assemblage structure in the region: habitat heterogeneity and proximity of coastal habitats. These natural drivers may counteract in some degree the lack of effectiveness of the MPAs and they will be discussed in more details below; however, improvements in MPA management are still essential for raising better ichthyofauna conditions (e.g., more abundant threatened species, higher biomass of critical functional groups).

Habitat heterogeneity is notably high in Los Canarreos region. In our study we recorded six different biotopes from 12 sites: crest, patch reef (shallow and deep), terrace, slope, and spur and groove. Caballero-Aragón et al. (2019) reported some of the largest values of reef rugosity (a measure of bottom complexity) in backreefs and forereefs of Los Canarreos compared to other sites around the Cuban archipelago. This heterogeneity promotes shelter and trophic resources for fishes (Gratwicke & Speight, 2005). Los Canarreos coral reefs are very close to other relevant coastal ecosystems, mainly mangroves and seagrass beds. The complex of interconnected mangrove-seagrass bed-coral reef ecosystems provides nursery areas that in turn enhance fish assemblages (Nagelkerken et al., 2002; Adams et al., 2006; Serafy et al., 2015). This may explain the relatively high taxonomic and functional diversity of fishes in the region. Even more, spawning aggregation sites occur for the snapper species Lutjanus griseus, Lutjanus analis, and Lutjanus cyanopterus in Cayo Avalos (Claro & Lindeman, 2003), which is relatively close to the study area.

Additionally, the proximity to many vegetated cays and the Zapata swamp (the largest swamp in the insular Caribbean) likely enhances the delivery of nutrients supporting high levels of primary production in the coral reefs. A recent study including 73 reefs around the Cuban archipelago (H. Caballero-Aragón, 2002, 2013, unpublished data) reports the highest concentration of particulate organic carbon and chlorophyll in the south of Zapata Swamp, relatively close to Los Canarreos. Other studies also reported nutrient enrichment in the coral reef system of Los Canarreos (Alcolado, Claro-Madruga & Martínez-Daranas, 2001; Reed et al., 2018; Weber et al., 2019). But important knowledge gaps remain for explaining inter-site variability of fish assemblages, which in turn weaken the success of management strategies. Examples of relevant knowledge gaps are the quantification of the habitat complexity and the effects of runoff from Zapata Swamp.

Our results suggest a need to improve the management/enforcement of Los Canarreos MPAs. All of the study area appears as protected/regulated, at least on paper. This “in-paper management” hides the current situation of this supposedly protected region and offer wide possibilities of illegal activities beyond of poaching, such as extraction of ornamental organisms such as black corals and some mollusk species. The fact that recorded sites are relatively far from human settlement represents a clear advantage for the MPA management, considering that human settlement tends to negatively affect the marine biodiversity (Mora, 2008). Therefore, the synergistic effects of natural and adequate management facts may preserve and even improve the ichthyofauna structure, the habitat condition, but also other processes that occur in this area, e.g., turtles nesting (Azanza et al., 2018). In addition, current touristic activities carried out in the area (SCUBA diving, and catch and release fisheries) could additionally benefit, since they depend on specific conditions including habitat mosaic and coral reef fishes.

Conclusions

The CCCR and CL MPAs lack significantly healthier coral reef fishes when compare to non-protected sites raising doubts about their effectiveness for conservation. Low abundance of threatened species and key functional groups, as well as low biomass, highlight the necessity of stronger management/enforcement in both MPAs. The current high taxonomic and functional diversities, and the occurrence of multispecies schools are likely remnants of healthy reef fish assemblages that used to inhabit the once pristine coral reefs of the region. The overarching effect of local environment (i.e., sites) points to the important role of habitat heterogeneity on fish assemblage structure. Favorable habitat features are likely enhancing fish assemblages and counteracting the effects of pouching derived from insufficient management.

Enhanced conservation of fish assemblages and the surrounding ecosystem in the eastern part of Los Canarreos archipelago would contribute to preserve the goods and services delivered by coral reefs and will be a significant contribution to the marine conservation of these ecosystems in Cuba and the Caribbean Sea. We recommend immediate actions within a strategy of precautionary management including, but not limited to, the appointment of staff for the administration of Cayo Largo MPA, frequent monitoring and effective enforcement. Improving the environmental quality of these coral reefs under effective management schemes will be instrumental for keep running touristic activities such as SCUBA diving and recreational fishing. Future studies should focus on the effects of habitats features (e.g., complexity, diversity) and surrounding coastal ecosystems (e.g., Zapata swamp) on coral reef fish assemblages.

Supplemental Information