**3. Results**

A total of 12,647 specimens were collected, representing 21 families, 43 genera, and 88 species (Supplementary Material, Table S1). For each quadrant, the number of species collected ranged from 13 to 38, and the number of individuals ranged from 36 to 705. The most diverse families were Alpheidae (21 species) and Porcellanidae (20 species). Ten families (47%) were represented by a single species (Supplementary Material, Table S1). The sample-based rarefaction showed that the sampling effort had an adequate representativity (79.6%) of the species richness expected by chance (Supplementary Material, Figure S1). The sampling effort ranged between 77.9% and 91.6% of representativity for all sites. The seasons showed 84.6% of representativity during the dry season and 85.7% in the wet season (Supplementary Material, Table S2). The representativity for year one was 79.5%, and for year two it was 88% (Supplementary Material, Table S2).

The most abundant species were *Trapezia corallina,* with 1720 individuals (13.6% of total abundance); *Trapezia bidentata,* with 1489; *Pachychelles biocellatus,* with 1028; *Petrolisthes haigae,* with 955; *Alpheus lottini,* with 820; *Petrolisthes hians,* with 619; and *Trapezia formosa,* with 579. Together, these species represented more than half of the collected specimens (Supplementary Material, Table S1). Of the total, 38 species (43% of the total) were represented by less than 10 individuals. Of these, 17 species had only 1 individual (singletons), and 4 had only 2 individuals (doubletons). Consequently, the contribution of singletons and

doubletons to the species richness was 23.8%. In addition, 14 species were collected in only 1 sample (uniques) and 7 in 2 samples (duplicates) (Supplementary Material, Table S1).

Individual-based rarefactions in pairwise comparisons showed that the species richness between sites was similar because their confidence intervals (95%) overlapped (Supplementary Material, Figure S2). An exception was Chamela and Carrizales, which had the highest and lowest number of species, respectively (Supplementary Material, Figure S2). The highest total species richness and abundance recorded over the sampling period were as follows: for Chamela, 69 species and 2371 individuals; for Cuastecomate-Punta Melaque, 64 species and 3266 individuals; for Carrizales, 58 species and 2752 individuals; and for Punto B, 68 species and 4258 individuals (Supplementary Material, Table S1). The total species richness was similar between years and between seasons (Supplementary Material, Figure S3). Year one showed 78 species and 5957 individuals, and year two showed 76 species and 6690 individuals. The wet season showed 79 species and 5276 individuals, and the dry season showed 73 species and 7361 individuals (Supplementary Material, Table S1). *Synalpheus arostris* and *Neogonodactylus pumilus* were recorded for the first time in the Mexican Pacific and showed a geographic extension of 3950 km to the north. Six other species were recorded for the first time in the Central Mexican Pacific: *Lophopanopeus frontalis*, *Daldorfia trigona*, *Pilumnus gonzalensis*, *Pilumnus reticulatus*, *Tumidotheres margarita,* and *Megalobrachium tuberculipes*. In Chamela, 50% of the species were collected for the first time, in Cuastecomate-Punta Melaque, 76%, and in Carrizales and Punto B, 63%.

The average taxonomic distinctness (Δ+) analysis at the site level showed that the Δ+ values for all the sites fell inside the probability funnel or within the 95% confidence intervals (*p* > 0.05). Chamela had the lowest Δ+ values despite having the greatest number of species (Figure 2). Punto B had the highest Δ+ values above the global Δ+ of the model. However, the Λ+ values for all sites fell within the probability funnel, indicating that the sampled sites were representative of the taxonomic diversity of the area. The seasons had different Δ+ values because the wet season fell within the probability funnel, but the dry season was outside the funnel (*p* < 0.05). The Λ+ values for the dry season were outside the funnel, so the taxonomic representativity during the dry season was lower than expected by chance (Figure 2). The Δ+ and Λ+ values between years were similar and fell inside the probability funnel (Figure 2).

The nMDS ordination showed that the taxonomic dissimilarity (Γ+) differed among the sites (Figure 3). The cluster analysis based on the SIMPROF procedure confirmed a group constituted by the southern sites (i.e., Carrizales and Punto B) and two separate entities (i.e., Chamela and Cuastecomate-Punta Melaque). This was also observed in the nMDS ordination. Carrizales and Punto B shared several species (e.g., *Pseudosquillisma adiastalta*, *Pomagnathus corallinus*, and *Synalpheus arostris*) and genera (e.g., *Trapezia*, *Liomera*, and *Pomaghnathus*), and they had almost the same families, except for the Pinnotheridae, which was only present in Punto B (and Cuastecomate-Punta Melaque). Conversely, Cuastecomate-Punta Melaque had a different crustacean fauna compared to the other sites and showed a mixture of taxa shared with Chamela and Punto B. Cuastecomate-Punta Melaque presented 34 genera and 17 families; these families were the same as Punto B, except for Panopeidae, which was found exclusively in this site, and Pseudosquillidae, which was absent. Chamela is the northernmost site and the most distant from the others. It was different because it had one superfamily (Parthenopoidea) not found elsewhere and two absent superfamilies (Eriphioidea and Pinnotheroidea). In Chamela, two families that were not found in other sites were collected (Parthenopidae and Lysmatidae), and four families (Panopeidae, Oziidae, Pseudosquillidae, and Pinnotheridae) were absent.

**Figure 2.** Average taxonomic distinctness analysis (Δ+) by site (**a**), season (**c**), and year (**e**); and its variation Λ+ for (**b**) site, (**d**) season, and (**f**) year. Codes: CH, Chamela; CT, Cuastecomate-Punta Melaque; CA, Carrizales; PB, Punto B.

**Figure 3.** Non-metric multidimensional scaling (nMDS) ordination shows the taxonomic dissimilarity of the crustacean diversity associated with *Pocillopora* corals among the studied sites in the CMP. Groups were separated as a function of the cluster analysis with an average group linking method and the similarity profile analysis (SIMPROF). Codes: CH, Chamela; CT, Cuastecomate-Punta Melaque; CA, Carrizales; PB, Punto B.

## **4. Discussion**

This study recorded most of the *Pocillopora* obligate symbiotic crustacean species reported by previous studies, including *Trapezia bidentata*, *T. corallina*, *T. digitalis*, *T. formosa*, *Alpheus lottini*, *Hapalocarcinus marsupialis,* and some species of *Synalpheus*. However, we did not find some species known to be associated with *Pocillopora*, such as *Fennera chacei*, *Alpheus sulcatus*, *Palaemonella holmesi*, *Stenorhynchus debilis*, *Thor algicola*, and *Petrolisthes galathinus*, which had been previously reported in the study area [24,25]. Nonetheless, we obtained two new records for the Mexican Pacific and six new records for the CMP (Supplementary Material, Table S1), increasing the known information regarding regional crustaceans.

Several species collected during this study, i.e., *Tumidotheres margarita*, *Typton* sp., and *Pontonia* sp., have been reported as endosymbionts of sponges, ascidians, or bivalves. These hosts are frequently associated with pocilloporid corals, and these decapods might be recognized as having a secondary association with pocilloporid corals. *Tumidotheres margarita* is an endosymbiont of the bivalves *Barbatia reevaena, Limaria pacifica,* and *Pinctada mazatlanica* [57], which are known as *Pocillopora*-associated mollusks in the Mexican Pacific [58]. *Typton tortugae* and *T*. *serratus* have been recorded as being associated with sponges living on corals [59]. In this study, some sponges were found to be associated with corals, and a similar association could exist in the cases of *T. hephaestus* and *T. granulosus*. Shrimps of the genus *Pontonia* are reported as obligate symbionts of the bivalves *Pinna* spp. and *P. mazatlanica* [60]. We assumed that the *Pontonia* specimens collected during this study were dislodged from their host during the collecting process or after the samples were preserved.

Our study increased the inventory of crustaceans associated with *Pocillopora* coral in the Mexican Pacific from 59 [20–24] to 88 species. Comparatively, in Huatulco, Oaxaca, a method similar to the one used here (0.25 m<sup>2</sup> quadrants) recorded 47 species of brachyuran crabs in pocilloporid corals [21]. In La Paz and Loreto Bay, Baja California Sur, 44 species of decapods were recorded [22]. Furthermore, a study covering almost the entire Mexican Pacific, from the Gulf of California to Oaxaca, recorded 36 crustacean species associated with pocilloporids [24]. The difference between the number of species reported herein and by Hernández et al. [24] may be a consequence of the visual census they performed. With this method, some close species are easily confused (e.g., *Synalpheus* spp., *Trapezia* spp., and *Alpheus* spp.) or overlooked (e.g., *Hapalocarcinus marsupialis*). The expected species richness estimated by the sample-based rarefaction was 20% higher than the observed richness due to the large number of rare species collected. Expected species richness is a good indicator of the potential species expected in the area. The sample-based rarefaction confirmed that the sampling effort was sufficient to elucidate the actual number of crustacean species associated with the *Pocillopora* coral in the CMP.

Decapod crustacean fauna associated with *Pocillopora* coral has been studied in many tropical and subtropical regions of the world's oceans. The species diversity recorded in this study is superior to the 36 species associated with *Pocillopora* off the Arabian coast in the Red Sea [61]. However, it is lower than the diversity reported from Oahu (Hawaii), where 127 species were found associated with *Pocillopora damicornis* [62], and then the 91 species reported more recently in 751 colonies of *P. meandrina*, also in Oahu [3]. For the northern Great Barrier Reef, Australia, 102 species were found in 50 colonies of *P. damicornis* [63]. It is important to mention that the obligate symbiotic composition observed in our study is similar to what has been reported for the Red Sea and the Great Barrier Reef, i.e., all three studies share the same brachyuran crabs (*Trapezia bidentata*, *T. digitalis*, and *Domecia hispida*) and caridean shrimps (*Alpheus lottini*, *Synalpheus charon*, and *Harpiliopsis depressa*).

A previous study indicated that the number of species present in coral ecosystems depends on the size of the coral colony [64]. The authors reported species richness ranging from 3 to 22 per colony (1500 cm<sup>3</sup> size) in the Gulf of Panama; in Costa Rica, 20 cm diameter colonies had 18 species [65]. Despite using quadrants of the same size, this study collected 13–18 species and 36–711 individuals per 0.25 m<sup>2</sup> of coral sample. These differences in abundance and richness are substantial and cannot only be attributed to colony size. To predict the species richness or abundance in colonies with stable conditions, some authors considered coral complexity (e.g., inter-branch space, penetration depth, and size of living space) [9], but in some cases, this factor was unable to explain the changes between different colonies [63]. For example, species such as the symbiotic *Trapezia* are not limited by coral complexity; they only need a healthy coral fragment for their survival [66]. Other characteristics, such as the percentage of live tissue and habitat degradation, could also influence the richness and abundance shifting. The species richness and abundance increase when the proportion of live coral tissue cover decreases [7,63]; this might happen because coral loss allows other species to move to new colonies. Moreover, coral mortality increases the abundance in single colonies [15], which may occur for two reasons: (1) the death of symbionts allows for other opportunistic species to move to more stable colonies, or (2) coral loss induces migrations of individuals looking for new space to live [7,9]. This situation could be happening in Punto B, where the coral colonies are isolated, fewer colonies are available, and the ecosystem is subject to anthropogenic pressure [18]. Symbiont loss does not seem to be a problem in Punto B because of the abundant obligate symbionts found in all samples.

The average taxonomic distinctness ( Δ+) varied between sites and seasons. The Δ+ values fell inside the 95% probability funnel, meaning they were a good representation of the taxonomic diversity of decapods and stomatopods associated with pocilloporid corals. However, Chamela had a lower Δ+ value despite having the greatest species richness among the four sites. This contrast occurred because Chamela featured the fewest supraspecific taxonomic hierarchies since many species belonged to the same families, i.e., Alpheidae (17 species) and Porcellanidae (18 species). In contrast, Punto B had the highest Δ+ value above the global Δ+ of the model and sustained almost the same species richness as Chamela. Punto B shared the taxonomic hierarchies with other sites and did not present any exclusive hierarchy.

Regarding the temporal variation of the taxonomic diversity, the Δ+ values fell outside the probability funnel in the dry season, meaning a relatively low taxonomic diversity change during this season; six genera (*Areopaguristes*, *Aniculus*, *Daldorfia*, *Bottoxanthodes*, *Pontonia*, and *Pseudosquillisma*), two families (Parthenopidae and Pseudosquillidae), and one superfamily (Parthenopoidea) were not recorded in this season. In contrast, the taxonomic diversity was better represented during the wet season, when the Parthenopoidea superfamily was present, portrayed by *Daldorfia trigona*, a species not collected in the dry season. Moreover, 15 species and 6 genera were exclusively collected during the wet season (Supplementary Material, Table S1). Years one and two had similar species richness and taxonomic structure. Likewise, both Δ+ values fell into the probability funnel close to the global Δ+ level, demonstrating that the studied years adequately represented the taxonomic diversity estimated by the global Δ+ model.

The nMDS ordination coupled with cluster analysis showed that Chamela had the highest taxonomic dissimilarities (Γ+) among the studied sites. Chamela—the northernmost site—was the most different with the highest taxonomic dissimilarity, the lowest Δ+, and the highest species richness. The Chamela samples contained one superfamily, two families, and four genera exclusive to this site, but several superfamilies, families, and genera present in the other sites were absent. It has been suggested that a low taxonomic distinctness can indicate a loss in the taxonomic diversity due to anthropogenic stress [30]. However, in this study, Punto B was the most anthropogenically affected site and displayed the highest Δ+ values. Despite moderate disturbances, symbiotic species tend to stay in their host for a long time [9,63]. Nevertheless, some symbionts (e.g., *Trapezia*) can migrate to other coral colonies in search of more suitable habitat [66,67]. Limited habitat availability makes them pile up in the colony, increasing species richness and abundance. This phenomenon could affect the Δ+ values in Punto B, increasing the values higher than the global Δ+ of the model. The low levels of taxonomic diversity in Chamela might be attributed to other variables, including the spatial process [68], benthic heterogeneity, habitat availability [69], or habitat

type [70]. In addition, it is important to remember that the variety of microhabitats is one of the main factors driving the diversity and abundance of coral-associated crustaceans [7].

In conclusion, the sampling effort in this study allowed for obtaining more than 70% of the expected species, indicating a good taxonomic representativity. The species richness and the taxonomic distinctness were within the expected values, despite being lower during the dry season. Most of the expected coral-obligated symbionts were collected, except for *Fennera chacei*, a small species frequently living in the coral base, which probably escaped during the collecting process. In contrast with the initial hypothesis, the sites with the most discontinuous coral cover and the largest human intervention did not have the lowest taxonomic distinctness (Punto B). However, as expected, the greatest abundance was observed in Punto B; this can be explained by the low coral availability, environmental variables, or anthropogenic stress. The present study should be complemented with α, γ, and β diversity analysis to assess the spatio–temporal differences in this particular species assemblage. It is also important to consider the influence of environmental variables, reef structural complexity, and human impact on the richness and abundance of these crustacean species, particularly in the obligate coral-symbiotic species. This study helped us to understand the crustacean assemblage associated with corals in the CMP and the spatio– temporal variations in their taxonomic diversity. Furthermore, it increased the taxonomic inventory of the coral-associated species in the studied region and the Mexican Pacific.

**Supplementary Materials:** The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/d14020072/s1, Table S1: Crustacean species list organized by families, Table S2: Sample-based rarefaction results, Figure S1: Sample-based rarefaction curves for the study area, Figure S2: Individual-based rarefaction curves between sites, Figure S3: Individualbased rarefaction curves between climatic seasons and sampling years.

**Author Contributions:** Conceptualization and methodology, A.A.-D., F.A.R.-Z. and M.A.-P.; formal analysis and investigation, A.A.-D. and F.A.R.-Z.; resources, F.A.R.-Z.; data curation, A.A.-D. and M.A.-P.; writing–original draft preparation, A.A.-D. and F.A.R.-Z.; writing—review and editing, F.A.R.-Z., M.A.-P., E.R.-J., M.d.C.E.-G., M.E.H. and O.V.-P.; project administration and funding acquisition, F.A.R.-Z. All authors have read and agreed to the published version of the manuscript.

**Funding:** A.A.-D. was funded by a doctoral fellowship (371662) from the Consejo Nacional de Ciencia y Tecnología (CONACYT). The scientific research project 257987 was funded by CB2015 from CONACYT.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** The authors confirm that the data supporting the findings of this study are available within the article and its Supplementary Material. Likewise, the data are available upon request from the corresponding author.

**Acknowledgments:** The authors would like to thank Sharix Rubio-Bueno and Karen A. Madrigal-González for their help in the fieldwork. We thank Enrique Godínez-Domínguez (CUCSUR-U. de G.) for his support with boats during the fieldwork.

**Conflicts of Interest:** The authors declare no conflict of interest.
