Environmental Drivers and Social Structure Features behind the Low Reproductive Success of Dusky Groupers Epinephelus marginatus (Lowe, 1834) in a Mediterranean Marine Protected Area

Abstract

Marine protected areas (MPAs) are considered key tools for achieving the sustainability of coastal fisheries by allowing habitat and species conservation by means of use regulations and active management. The effects of protection are more visible in slow-growing, high site fidelity and late-maturing species such as dusky grouper, a flagship species for all the fishermen and divers in the Mediterranean Sea. We studied the reproductive behavior of dusky grouper at Tabarca Island Marine Protected Area (TIMPA) in order to: (a) describe its temporal pattern; (b) evaluate the relationships among reproductive behavior and environmental factors and (c) verify the effects of social structure on reproductive activity. We observed all the behavior patterns previously reported in other studies besides the final spawn event. Our results evidenced that the monthly average density of dusky groupers tended to increase when temperature rose close to the seabed and without defined temporal limits for reproductive behavior. Among the environmental factors considered, visibility exhibited a very strong effect on the reproductive behavior of dusky grouper, interacting significantly with seabed temperature and thermocline. Regarding the effects of social structure, reproductive activity was positively correlated with the presence of large-sized individuals. We have discussed the differential facts of the reproductive activity of dusky groupers observed at TIMPA relating the results to the biogeographical location, the patterns of the environmental drivers (such as the seabed temperature per day, the thermocline, the visibility, the stream force, the photoperiod and the divers per day) and the structure of the habitat. These negative results contrast with the existing knowledge and generate new working hypotheses useful to improve the efficiency of MPAs and the sustainability of coastal fisheries.

1. Introduction

MPAs are considered key tools for achieving the sustainability of the seas [1] in order to maintain ecosystemic services to the human population and are widely proposed as essential tools for marine resources management [2]. Through use regulations, MPAs allow for habitat and species conservation [3] and restore lost ecosystem processes [4]. The effect of protection is variable among species [5], being most visible in slow-growing, high site fidelity and mature late species that are more sensitive to exploitation, such as groupers in the case of Mediterranean MPAs [6]. Reproductive aggregations of groupers are very important socially and economically [7] and MPAs are considered essential to ensure their survival and functionality. The limitations imposed by the management of MPAs generate a decrease in fishing mortality, which is much more evident in these slow-growing and late-maturing species [8]. Consequently, the individuals of these species reach higher ages and, with it, higher potential fertility [9]. After a period of protection, while the populations recover within the boundaries of the MPA, the increases in potential fertility serve to repopulate individuals inside of MPAs [10]. Afterward, due to density-dependent factors [11], new individuals will exit the MPA via the exportation of adult individuals [12] or the spillover of eggs, larvae and juveniles [13], interconnecting populations at larger scales [14]. In these circumstances, the study of reproductive behavior and the verification of spawning is essential for assessing population dynamics and the effect of the protection of the MPA on these threatened target species and to verify that the proposed objectives for the AMP are achieved (among others, the recovery of the population structure and the ability to repopulate the areas surrounding the MPA to sustain coastal fisheries). The dynamics of these processes can change according to the environmental factors of a specific area at scales of tens of kilometers and the population dynamics of the species in that area [11], which makes it necessary to study reproductive behavior at these local scales.

As commented above, among the most important species, groupers (Serranidae, Epinephelinae) are of significant economic and recreational importance [15]. The dusky grouper Epinephelus marginatus (Lowe, 1834) is one of the most commercial and recreationally important species in the Mediterranean coastal areas [16], assessed as overexploited [17]. For this reason, it is included as endangered species in Annex 3 of the Berne Convention and Annex 3 of the protocol for specially protected areas and Mediterranean biodiversity of the Barcelona Convention, with a mandate to implement effective management measures to ensure its conservation. In MPAs, E. marginatus may be adversely affected by some features of the species’ inherent biology working synergistically with local fisheries. This is a sequential hermaphroditic protogynous species, reaching sexual maturity at five years of age. When their sizes are small and medium, individuals tend to look for new territories outside the reserve due to density-dependent factors linked with the territorial behavior of the species [10], where they are frequently caught by artisanal fleets. These captures, according to Sadovy de Mitcheson et al. [7], can eventually end up skewing the population of large-sized individuals or males. Given that MPAs were designed to assure the viability and sustainable use of these species, it is important to assess and verify their effective reproduction into the limits of the managed area.

The reproductive behavior of grouper has been studied in different Mediterranean locations. Zabala et al. [18] reported the first systematic direct description of the reproductive behavior of grouper, its relationship with environmental factors and its temporal development in the Marine Protected Area of Medes Islands (Girona, Spain, NW Mediterranean), as well as the social context of the species in which reproduction was observed [19]. The direct observation of reproductive aggregations has also been repeated in different protected localities in southern France and Corsica [20,21] and an unprotected area in northern Corsica [22]. Indirectly, dusky grouper reproduction has been verified in southern France through the presence of eggs and larvae [23], in Balearic Islands [24], in southern Italy [25] and in waters from Turkey [26], both using cytological studies of the gonads. Other studies report evidence of dusky grouper reproduction in southern France [27] and northeast of Sardinia [28] based on telemetry recordings of movements and provided correlations between reproductive activity and different environmental factors to try to predict the temporal occurrence of spawning in any Mediterranean locality [29]. In all cases, reproductive behavior and gonadal development are observed from late spring to late summer, with peaks of reproductive activity between July and August and no evidence of reproductive activity outside of this time.

Intending to start the study of the reproductive activity of dusky grouper in the Tabarca Island Marine Protected Area (TIMPA) (Alicante, Spain), we organized a sampling based on the existing bibliography in a location similar to those described in the works mentioned above and situated in the restricted-use zone of TIMPA. We initiated direct observations of the groupers to describe their reproductive behavior and verify the spawning, relating it with relevant hydrographic features which ensure the survival of the early life stages and cause potentially high predictability of reproductive events [30]. Preliminary observations confirm the aggregation of dusky groupers, with the social behavior features similar described by prior literature [18,19,20,29]. A larger medium-sized proportion than large-sized, territory patrol behavior by large males and the complete range of interactive corporal and color patterns described by the cited studies was observed (Lozano-Quijada, unpublished data). Despite this, we neither observed the mating nor the spawning. To verify the length of the reproduction period, we sampled monthly the locality over a year (2016), observing that the reproductive behavior also occurred outside the summer season reported in the studies mentioned above, which led to thinking about the existence of particularities that could be interfering in the development of the reproductive activity of the dusky grouper in the Marine Reserve of Tabarca. For this reason, we decided to extend the study of dusky grouper reproductive activity in the Marine Reserve of Tabarca in order to document this process in detail and with three specific objectives: (a) describe the temporal pattern of dusky grouper reproductive behavior, (b) evaluate the relationship of reproductive behavior with environmental drivers and (c) verify the effect of social structure on reproductive activity.

2. Materials and Methods

2.1. Sampling Area and Experimental Design

The social and environmental effects on the reproductive behavior of E. marginatus were tested by collecting environmental variables and noting its social structure and behavior at a grouper hotspot (38°09.222′ N; 00°25.847′ W) placed into the regulated-use zone of TIMPA from June 2017 to September 2019 (Figure 1). This Marine Protected Area is part of the Spanish 12 Marine Reserves of Fishing Interest. It was created and co-managed by the Ministry of Fishing of Spain (Secretaría General de Pesca) and the Conselleria de Agricultura y Pesca (Regional Government), of which TIMPA was the first established in 1986 (https://www.mapa.gob.es/es/pesca/temas/proteccion-recursos-pesqueros/reservas-marinas-de-espana/default.aspx, accessed on 14 March 2022). Its main aim is to maintain the sustainability of artisanal fishing around the marine protected area and regenerate the targeted stocks, as well as to conserve the existing marine biodiversity. This fishing vocation justifies the authorization of some regulated low-impact gears into the marine protected area [31].

The hotspot seabed consisted of flat rock bottom and Posidonia oceanica (L.) Delile, 1813 meadows at 24 m depth (Figure 2a). It was located eight nautical miles from Alicante submarine canyon, which generates particular oceanographic conditions at certain times of the year. Here, recreational diving was regulated and only a maximum of 36 divers was allowed per day. Fishing was only allowed for professional pelagic trolling, so the biodiversity and fish abundance were very high [32]. The grouper hotspot was composed of 7 sub-areas or sites, each one with similar landscape features, except a small promontory emerging in the middle of flat rocky-Posidonia meadow (Figure 2c), 4–5 m high and oval-shaped, measuring about 35 m of major diameter and 20 m of minor diameter (Figure 2c). This promontory served as a meeting point for the dusky groupers which allowed us to discover the reproductive aggregations. The total area surface was about 6000 m² (Figure 2d). We recorded sexual behaviors and social structure by underwater visual census, following the stationary point census method [33] in an established survey through the seven sites (Figure 2b) for 40–45 min. The diving depth was about −15 m and the sampling area comprised a circle of 10 m radius centered at the sampler. We performed fifty-eight surveys, randomly distributed over the studied period, during the sixty minutes after sunset, when water conditions were optimal (visibility more than 10 m and swell less than 0.5 m height) to avoid bias due to environmental constraints [33]. There are no data for some months due to bad weather conditions or very low visibility.

During each census, we took note of:

• Environmental drivers:
  • The seabed temperature per day (°C). Temperature data were recorded each hour by a HOBO Pendant UA 002-64 data logger placed at 70 cm up the seabed and averaged per day;
  • The thermocline, estimated by sea surface–bed temperature difference. Large differences were linked to a stable thermocline, whereas small differences to a weak or nonexistent thermocline;
  • The visibility was established qualitatively by the sampler as bad (<10 m), moderate (from 10 to 20 m) or good visibility (>30 m);
  • The stream force over the seabed was classified the same way as the previous variable: (1) no stream, (2) moderate, (3) strong and (4) very strong stream;
  • The photoperiod or day length in hours at the hotspot position each survey-day. These data were consulted from the National Meteorological Agency (AEMET);
  • The number of recreational divers who visited the hotspot per survey-day which was obtained from the MPA surveillance enterprise.
• Social structure and behavior variables:
  • The density or specimen number per survey;
  • The size class. Groupers were classified into three size classes, smalls (30 to 50 cm), mediums (50 to 80 cm) and larges (>80 cm). The size classification is strongly linked to sexual maturation [24];
  • The group size. We classified the size group where the sighted grouper was included in (1) an alone specimen, (2) small (2 to 5 specimens), (3) medium (6 to 15 specimens) and (4) large (16 to 30 specimens) groups;
  • The color pattern. We classified color patterns into (1) standard mottled, (2) defending or attack and (3) reproductive. This classification was summarized from the eight color patterns reported in the literature [19,20] in order to achieve a more effective response for our aims;
  • The behavior. Its classification was summarized, as the pattern color, from those reported in the literature cited. It consisted of (1) non-sexual behavior, (2) pursuit or escape, (3) territory patrol, (4) female showing reproductive behavior and (5) male in courtship;
  • The reproductive status. This qualitative variable indicated the sexual state of the specimen when it was sighted and only admitted two states: (1) reproductive or (2) non-reproductive status. It was determined by the behavior and the color pattern;
  • The position. Specimen position was determined as (1) swimming, (2) motionless, resting at the water column, (3) hidden or (4) perched on the seabed;
  • The belly. A swollen belly was representative of a female with matured eggs; we noticed if it was (1) swollen or (2) flat.

2.2. Statistical Analysis

The environmental and social factors that affected reproductive status were tested using Generalized Linear Mixed Models (GLMMs) with a binomial family distribution and a logit link function [34]. This statistical approach allowed us to test the fixed and random effects on our binary response variable, the reproductive status, by an unbalanced design. The estimation of parameters was performed by Laplace approximation, running the glmmADMB software package [35,36] on R-software [37]. Laplace approximation is a more accurate estimate method than penalized quasi-likelihood, but it is slower and less flexible [38]. Due to the large number of fixed effects tested, we decided on segregating them into two different GLMM models, the environmental and the social models, in order to obtain results more easily interpretable [39]. Fixed factors were sampled at the same seven sites for three years, so we chose sites as a random effect in both models, to account for the lack of independence or pseudoreplication associated with the collection of multiple measurements from the same sites over the time. The years were discarded as a grouping variable due to its limited number of levels, only three, and their low representativeness because they were sequentially sampled [39,40].

Environmental and social models were optimized with a backward selection procedure of fixed factors in which we removed non-significant terms from the models, one by one, in decreasing order of their p-values [41]. The goodness of variance fit was tested using Akaike’s Information Criterion (AIC). In the environmental model, a triple-crossed effect of average seabed temperature, thermocline and visibility was included to obtain a better model fit. In addition, all the continuous variables were standardized to improve the interpretability of regression coefficients and to make main effects biologically interpretable even when they are involved in interactions [42]. The optimal models were:

$$Y_{ij}Bin(1,p_{ij}).$$

(1)

The specific environmental model was:

$$logit\left(p_{ij}\right)=\alpha +{\beta }_{1}T+{\beta }_{2}The+{\beta }_{3}V+{\beta }_{4}T\ast The+{\beta }_{5}T\ast V+{\beta }_{6}The\ast V+{\beta }_{7}T\ast V\ast The+{\beta }_{7}Ph+a_i;$$

(2)

where T = seabed temperature, The = thermocline, V = visibility, Ph = photoperiod and a = sites. The specific social model was:

$$logit\left(p_{ij}\right)=\alpha +{\beta }_{1}SC+{\beta }_{2}P+{\beta }_{3}SG+a_i;$$

(3)

where SC = size class, P = position on the seabed and SG = size group;

$$a_{i}N(0,{\sigma }_a^2);$$

(4)

where $Y_{ij}$ was 1 if a grouper j on a site i (from site $i=1$ to 7) showed a reproductive status and 0 otherwise. The random site intercept $a_i$ was assumed to be normally distributed with mean 0 and variance ${\sigma }_a^2$. If random effects variance was small, then the contribution of sites was also small and they would have similar logistic curves [29].

3. Results

Throughout the study time, we observed up to six dominant males exhibiting a silver-streaked livery, or “livery 6”, according to Zabala’s description [19]), patrolling simultaneously in their respective areas (Figure 2b), without overlapping and moving right above the ground in a very active way around a reference point. Patrol areas were approximately 35 m in diameter, similar to those described by other authors. In contrast to what has been reported by other studies, dusky groupers patrolling at TIMPA were always observed at the same depth, below the summer thermocline and most frequently around the small promontory that emerged in the middle of the flat rocky Posidonia bottom. We observed many courtship events between males and females, in which the male performed the typical movements described by Zabala et al. [19] (flappings and tiltings) while the female initially maintained the approach with the male. However, the female always ended up retreating or even fleeing pursuit by the male and did not spawn.

3.1. Temporal Pattern of Dusky Grouper Reproductive Behavior

The monthly average density of groupers in the hotspot tended to diminish when seabed temperature reached minimums and to increase when the sea water was warmer. Its variability was higher during the warmer months, among years and among the warmer months of the same year (Figure 3). The proportion of small specimens (considered females) increased in warm months. Conversely, in cold months, the alone and large and medium specimens predominated in the hotspot, whereas the smalls disappeared (or at least were much less conspicuous) (Figure 4). The first reproductive signals began in spring and they lasted until the end of autumn. There were no defined limits for the reproductive period. Over an annual scale, the earlier reproductive behaviors were detected in March (2019) whereas the latter ones were in December (2018), although it could be confused with territorial displays. The reproductive color patterns and sexual behaviors were more usual in warmer months, although surprisingly, we did not record any spawning in all the study time at the hotspot (Figure 5 and Figure 6).

3.2. Relationship of Reproductive Behavior with Environmental Drivers

The reproductive status in E. marginatus is strongly affected by seabed temperatures, thermocline and visibility and moderately by the photoperiod (

Table 1. Coefficients of fixed effects with their confidence intervals and hypothesis tests estimated by environmental Generalized Linear Mixed Models (GLMM). The standard error and the exponential were calculated for each coefficient estimated. In addition, Akaike’s Information Criterion (AIC) and standard deviation were calculated for the random effect (sites).

Variables	Estimate	Std. Error	Exp. Estimate	Lower 95	Upper 95	p-Value
Intercept	−0.158	0.305	0.853	0.470	1.550	0.603
S(T)	0.794	0.159	2.213	1.622	3.021	<0.001
S(The)	1.045	0.170	2.844	2.037	3.971	<0.001
Moderate V	−0.127	0.206	0.881	0.589	1.319	0.538
Good V	0.190	0.0289	1.209	0.687	2.130	0.510
S(Ph)	0.214	0.117	1.239	0.985	1.558	0.067
S(T):S(The)	0.236	0.151	1.267	0.942	1.703	0.118
S(T):Moderate V	−1.338	0.271	0.262	0.154	0.446	<0.001
S(T):Good V	1.072	0.408	2.920	1.313	6.492	0.009
S(The):Moderate V	−1.049	0.244	0.350	0.217	0.566	<0.001
S(The):Good V	−0.435	0.521	0.647	0.233	1.798	0.404
S(T):Scale(The):Moderate V	−0.672	0.310	0.511	0.278	0.938	0.03
S(T):Scale(The):Good V	1.351	0.534	3.861	1.356	10.995	0.011
AIC	1205.394
Sd. sites (Intercept)	0.725

T = seabed temperature, The = thermocline, V = visibility, Ph = photoperiod. If the fixed effect is preceded by “S”, then it was standardized.

, Figure 7). High seabed temperature, strong thermocline and visibility interact positively if visibility is high but negatively if it gets low. In addition, the confidence intervals (CI) are wide when the visibility is good (1.356 to 10.995) and narrow when it is moderate (0.278 to 0.938). The good visibility promotes reproductive behavior; however, it does not happen with the same intensity in all sites. There are large differences among sites, as confidence intervals indicate. However, when the reproductive status is inhibited by high or moderate turbidity, it is similarly inhibited at all sites. This fact is also corroborated by its narrow standard error (std error 0.31), which indicates a scarce behavioral variation among subjects in high or moderate turbidity. A comparison of exponential estimates between both triple interactions suggests that low turbidity effect on reproductive status was higher than the high or moderate turbidity effect by 7.5-fold (the exponential estimate of the triple interaction was 3.861 vs. 0.511 ). Photoperiod was not a significant fixed effect, but its p-value was very close to 95% confidence (p-value = 0.067) and its trend was closely linked with reproductive status dynamics around the surveyed years (Figure 7), so we considered that it was playing an overlapping effect related with turbidity to the triple interaction exposed above. On the other hand, although our direct observations seemed to corroborate a negative effect of the current intensity on the reproductive behavior of dusky groupers, it had no significant statistical effects and was discarded in the environmental model optimization process.

3.3. Effect of Social Structure on Reproductive Activity

High reproductive activity was strongly linked to the large-size class (exponential estimate 3.82;

Table 2. Coefficients of fixed effects with their confidence intervals and hypothesis tests estimated by the social Generalized Linear Mixed Models (GLMM). The standard error and the exponential were calculated for each coefficient estimated. In addition, Akaike’s Information Criterion (AIC) and standard deviation were calculated for the random effect (sites).

Variables	Estimate	Std. Error	Exp. Estimate	Lower 95	Upper 95	p-Value
Intercept	−0.938	0.277	0.391	0.288	0.673	0.001
Medium SC	0.524	0.358	1.690	0.837	3.410	0.143
Large SC	1.360	0.240	3.897	2.433	6.243	<0.001
MotionlessP	−0.276	0.170	0.759	0.543	1.060	0.106
HiddenP	−2.659	0.611	0.070	0.021	0.232	<0.001
PerchedP	−3.143	0.471	0.043	0.017	0.109	<0.001
Small SG	−1.078	0.213	0.340	0.224	0.517	<0.001
Medium SG	−2.011	0.498	0.134	0.050	0.355	<0.001
Large SG	−13.805	479.98	0	0	∞	0.977
AIC	1131.382
Sd. sites (Intercept)	0.404

SC = size class, SG = size group, P = position on the seabed. The whole large group showed specimens in non-reproductive status; for this reason, its estimated coefficient is so large and dispersed.

). The confidence interval of this effect was wide (2.433 to 6.243) and reflected the different role of sites in the sexual behavior of dusky groupers, whereas the small variations in standard error (0.24) reflected the consistency of behavior within each site. When they were hidden or perched on the seabed, this evidenced a significant but weak negative effect on reproductive activity, as the small exponential estimates suggested (0.07 and 0.043, respectively). During the warm season, it was usually found that the groupers perched on the seabed, showing a characteristic postreproductive light color pattern, whereas in the cold season, they were hidden in their caves. The belly shape and the frequentation had no significant effects on reproductive status and they were rejected in the social model backward selection procedure.

Groupers tended to congregate during warm-weather periods, although large groups were rarely observed (Figure 8), and when they were found, no individual showed reproductive behavior. Our model detected a negative effect on sexual activity when groupers were kept in small and medium sizes (exp. estimates 0.34 and 0.134, respectively). These results reflected the female groups resting in warm waters over the thermocline (up the rocky promontory).

4. Discussion

Our observations on patrolling territory size, male liveries and patterns of behavior coincide with those described by other authors [18,19,29], although they differ in the depth range in which groupers carry out their reproductive activity. The results evidenced a reproductive behavior of the dusky grouper in the TIMPA marked by seasonality, although with important intra- and interannual variations. Reproductive behavior was more intense during the warm months but closely linked to the bottom temperature, the structure of the thermocline and visibility, with the latter having a very important regulatory role on the other two environmental drivers. Likewise, reproductive activity is closely linked to the frequency of large patrolling individuals. With this, we confirm the existence of reproductive aggregations of dusky grouper in the Tabarca Marine Reserve, although some characteristics (number of individuals, size composition and behavior) are different to those described in other studies [15,25] and without observing a complete spawning in all the time that the sampling lasted.

The first differential fact to highlight in TIMPA is the duration of the reproductive period. In other localities [18,20,22], reproduction lasted markedly between the months of July and August, mainly. In our study, reproductive behavior was observed throughout practically the entire year, although with high interannual variability. The fact, on the one hand, of being in an area of the Mediterranean with higher seawater temperatures [43] could explain the lengthening of the breeding season, as occurs in southern Mediterranean locations [25] or in more tropical environments [44], where individuals with mature gonads are found outside the summer period. The intensity of the reproductive process has been closely linked to the presence of dominant male individuals, with highly accentuated color patterns, patrol movement and defense of the territory, as described by other authors [18].

On the other hand, the oceanographic conditions marked by the bottom temperature, the thermocline and the visibility explain an important part of the variability of our observations. The reproduction of the dusky grouper occurs when the surface temperature is above the range of 25–26 °C, as reported in other studies [18,29,45], which is deemed optimal for the development of gonads and eggs in some grouper species [46,47]. Visibility results as the environmental driver that modulates the rest of the significant environmental factors and determines the reproductive behavior of the grouper. In the study area, visibility decreases mainly due to the contribution of particulate matter from the Bay of Santa Pola, from the west, and/or the Bay of Alicante, from the north, generated by the winds of the third and fourth quadrants (https://www.puertos.es/es-es/oceanografia/Paginas/portus.aspx, accessed on 12 march 2022), acting as an environmental perturbation which blocks the progression of the reproductive process. It would explain the attenuation of reproductive behavior within the same season, during July and August 2018, caused by the strong thermocline, which promoted the decrease in visibility by particulate matter suspended on the upper thermocline layer and the cool water in the lower layer.

High visibility (equal to or greater than 20 m) would accentuate reproductive behavior when the bottom temperature was high and the thermocline was well-formed; low visibility (between 10 and 15 m) would attenuate reproductive behavior. The high visibility values in which a high reproductive behavior has been confirmed are the same as those reported in other MPAs (e.g., Medes Islands; [18]). This attenuation of reproductive behavior due to low visibility was observed for all the sampling sites simultaneously and on all the individuals, which shows its global regulatory role over the total studied area. The appearance of particulate matter in suspension affects several processes in the biology of fish [48] including reproduction [49] and, given our results, it may be an important issue that determines the role of AMPs as mechanisms to favor the reproduction of sensitive species such as dusky grouper.

The second differential fact is the absence of complete spawning at the end of the reproductive interaction. We never observed the spawning of dusky grouper after several years of direct observations, not even after the systematized quantitative data collection between 2017 and 2019 plus subsequent recordings to date (February 2022). The lower total density of individuals and the number of female individuals per group in TIMPA could explain this specific circumstance. The size structure of dusky grouper at TIMPA is also different, with a lower proportion of medium and big individuals. For example, while in our study area the density is over 40 individuals per transect and the most frequent groups are those formed by 2–5 specimens, in the Medes Islands there were reported 120 individuals per transect, with groups formed by more than ten specimens. In this locality, the greater heterogeneity of the habitat, dominated by large rocky blocks and many hollows, would generate an aggregating effect on the dusky groupers, favoring higher densities of individuals and providing more refuge. During the cold season, the dusky groupers occupy these available refuges, from which they do not leave (M. Zabala, pers. commun.), remaining within the marine protected area. TIMPA does not have a habitat with such spatial complexity nor such an amount of refuge, which could affect the density of individuals, their size structure and their permanence within the marine protected area during the cold season, explaining the differences found in TIMPA with respect to the results reported in the commented other studies. On the other hand, the proportion of individuals in each size class, with a lower number of medium and large females than that described in other MPAs, could be attenuating the reproductive process due to the lack of truly fertile females in those existing sizes [50].

On the other hand, around TIMPA is an area subject to intense commercial fishing pressure. Groupers between 40 and 60 cm (per. obs.) are frequently caught by the artisanal fleet outside the limits of the reserve as a result of the export of biomass [10]. The small size of TIMPA, together with the intense fishing activity in its surroundings, favors the export of small- and medium-sized individuals [51]. These facts would prevent higher densities and a suitable size structure for reproduction from being reached within the reserve. It would also be favored by the continuity in the area of the Posidonia oceanica meadow, which would accentuate the biomass gradient towards the outside [32]. All these factors together would be generating an edge effect [52] inside TIMPA that would be undermining its potential to recover threatened species such as grouper and favor their spawning.

A third issue that would be conditioning the success of reproduction in TIMPA would be the structure of the habitat itself. In the MPAs where the spawning of dusky grouper has been described, the rocky habitat extends from −8 m to −30 m, in a maximum direct length of about 20 m. In this habitat structure, the dominant males move from the deepest part to the shallowest area to interact with the female and spawn [27,28] when the surface temperature is maximum (more than 26 °C) and the meteorological conditions are anticyclonic [29]. In TIMPA, the habitat is a massive flat rock with hardly any loose blocks present, excepting the small promontory emerging in the middle of the rocky-Posidonia meadow where dusky groupers aggregate more frequently, extending between −24 and −27 m in depth, most of the time below the summer thermocline. This habitat structure could be constraining social reproductive behavior. Adequate habitat structure may be more important than has been described so far, determining the reproduction of dusky grouper. In the Cabo de Palos Marine Reserve, 63 km south of TIMPA, with similar fishing and regulating circumstances, but with a habitat configuration such as that described in localities where the complete reproduction of the grouper has been confirmed, spawning was observed repeatedly during the summer of 2021, a few days after beginning a monitoring programme on the reproduction of dusky grouper (Luís González and J.A. García-Charton, per. com.). Therefore, in the combination of factors necessary for the effective reproduction of the grouper to occur, the suitable structure of the rocky habitat could be one of the most important. More research in this direction is necessary.

Given the multitude of factors that can condition the effectiveness and sustainability of an MPA, our study on dusky grouper reproductive aggregations in TIMPA, with negative results, contributes to contrasting the positive results reported by other studies and allows progress in determining the role of difficult-to-assess driving factors, such as disturbance from external physical drivers, management issues or habitat structure. The overlapping of ecological causes with different space-time dynamics [53] can lead the MPA to fail. The uncertainty associated with these circumstances and the probability of failure can be mitigated through the application of adaptive management [54] based on monitoring results [55] (Nikols et al., 2019) such as those exposed here to find global application patterns to improve the performance of MPAs and achieve their sustainable goals.

5. Conclusions

Our results have evidenced that the effective spawning of dusky grouper has not been carried out, despite verifying the aggregation and reproductive behaviors (lots of courtships, patrolling, male color pattern 6, bellies of females, etc.). Increased turbidity generated by oceanographic processes external to TIMPA and other overlapping ecological processes may be the reasons that spawning is not lastly carried out effectively in TIMPA. Our work, even showing negative results, provides content to contrast the existing knowledge on dusky grouper reproduction and generate new working hypotheses to improve the efficiency of MPAs and the sustainability of coastal fisheries.