Richness and large-scale distribution of marine benthic caridean shrimps ( Decapoda : Caridea ) from the Eastern Tropical Pacific

Based on an updated checklist of the benthic caridean shrimp fauna of the Eastern Tropical Pacific (ETP), we analyzed their large-scale species richness distribution patterns using a grid approximation. Caridean fauna is composed of 183 species belonging to 67 genera and 18 families. Alpheidae, Palaemonidae and Thoridae contributed largely to species richness. Alpheus and Synalpheus were the most diverse genera. Most species (24%) have a narrow distribution inhabiting a single grid, and as many as 135 are restricted to < 5 grids. Mexico has the largest richness (130) besides Ecuador (98) and Panama (75). Richness among grids varies largely; overall, it is larger around Galapagos and Panama, Colombia, Costa Rica and the Gulf of California. There was a meaningful correlation between a proxy of sampling intensity and richness, suggesting that along with habitat heterogeneity or large-scale evolutionary processes currently invoked as richness drivers, the pattern can also be related to differences in research efforts. Exponential fitting and Chao2 index suggest that the caridean faunal inventory from the ETP is still far from being complete. Overall, the results call for a more thoughtful and systematic sampling scheme in order to get information from poorly sampled environments and areas in the ETP.


Caridean shrimp from the Eastern Pacific
Nauplius, 26: e2018035 inTroduCTion Shrimps included in the infraorder Caridea are the most diverse group of decapod shrimps (Gorny, 1999;De Grave and Fransen, 2011).Caridean shrimps are mostly benthic organisms inhabiting muddy to rocky bottoms from a wide diversity of habitats (e.g., tidal pools, seaweeds), either free or associated to other marine organisms (Wicksten, 1983;Bauer, 2004;Guzmán, 2008).They can be found from intertidal to hydrothermal vents, and in freshwater, estuarine and marine habitats worldwide, from the equator to the poles but mainly inhabiting tropical waters (Carvacho and Ríos, 1982;Martin, 2003;Anker et al., 2006a).
Caridean shrimp studies in the Eastern Tropical Pacific (ETP) started as early as 1869 in Panama (Smith, 1869) and studies in the Pacific coast of Mexico and Ecuador contributed steadily via the discrete addition of new species to the regional fauna (Faxon, 1893;1896).This early trend was occasionally disrupted by large-scale works where the caridean taxonomy of large areas (Coutière, 1909) or entire supraspecific groups (Holthuis, 1951;1952;Kim and Abele, 1988) were addressed.Recently, as a result of studies conducted by local taxonomists (Abele and Kim, 1989;Wicksten, 1991;Li and Poupin, 2009) and ecologists (Hernández et al., 2013), there has been an increase in the knowledge of the caridean shrimp fauna from the ETP.
The most recent checklist of the benthic caridean shrimp fauna from the ETP was published by Wicksten and Hendrickx (2003).Since then, species descriptions (Wicksten and Martin, 2004;Hermoso-Salazar and Álvarez, 2005;Anker et al., 2006a;2007;2009;Hermoso-Salazar and Hendrickx, 2006;Li, 2006;Marin and Anker, 2008;Li and Poupin, 2009;Hendrickx, 2010), geographic range extension (Hermoso-Salazar and Hendrickx, 2005), and changes in taxonomy and systematics (Marin, 2009;Bracken et al., 2010) have substantially modified the previously known information concerning caridean shrimp faunal inventory and its distribution patterns in the area.Current distribution patterns of ETP caridean shrimps were known after studies focused on more inclusive groups such as macroinvertebrates, crustaceans or decapods (Arntz et al., 1999;Blanchette et al., 2008) and at large-scale levels (i.e., Province) (Boschi, 2000;Wicksten and Hendrickx, 2003).For example, Boschi (2000) indicates that the so called Panamic province hosts the largest number of both planktonic and benthonic caridean species (161), followed by the Galapagos (67 species) and Cortes (50 species) provinces.Additionally, the same general pattern emerges when only benthonic caridean species are considered (Wicksten and Hendrickx, 2003).While there is an overall agreement among studies regarding richness patterns, the previously used spatial scale precludes direct comparisons among areas which obscure intraregional species richness patterns.Beside taxonomic bias, databases resulting from field exploration projects and/or previously published information allow for large-scale analyses that evaluate current species richness and distribution scenarios (Costello et al., 2010).In this regard, there is a need to integrate previously published findings such as spatially explicit checklist, to gain perspective on what is known and what are the major scientific gaps regarding benthic caridean shrimp species richness and distribution in the ETP.
In this paper, species composition and distribution of the benthic caridean shrimp fauna of the Eastern Tropical Pacific are analyzed.An updated (up to 2017) checklist of the benthic caridean shrimp fauna from the ETP, at a 1° of latitude by 1° longitude, was assembled and further analyzed to address a set of interrelated questions.What is the richness and composition of the ETP benthic caridean shrimp fauna?Has the accumulative historic species record been homogeneous over the entire region or do discrepancies exist among areas?Does species richness show any distribution pattern in the area?Are ETP benthic caridean shrimp species widely or narrowly distributed in the area?Finally, how complete is the benthic caridean shrimp fauna inventory?

MaTerial and MeThods
We compiled a checklist of the marine benthic caridean shrimps that have been collected and/or recorded in the coastal and shelf area of the Eastern Tropical Pacific.A database was generated from primary references that have been formally published as papers, monographs and books (mainly, but not restricted to Kingsley, 1878;Faxon, 1893;Rathbun, 1904;Coutière, 1909;Chace, 1937;1962;Holthuis, 1951;1952 1975; Abele and Patton, 1976;Wicksten, 1983;1991;2000;Kim and Abele, 1988;Abele and Kim, 1989;Lemaitre and Álvarez-León, 1992;Wicksten and Hendrickx, 1992;2003;Burukovsky, 2001; complete reference list is included in Tab. 1).In order to minimize taxonomic bias, only valid species were included, hence all taxa names were crosschecked in the World Register of Marine Species (WoRMS) and Integrated Taxonomic Information System (ITIS).Systematic arrangement of benthic caridean shrimps follows De Grave and Fransen (2011).The taxa (genera and species) were arranged alphabetically (Tab.1).After checking for species validity, historic species accumulation curves (species added by year of publication) were constructed for each country and for the entire ETP.Further, the data was fitted with an exponential function (larger r 2 ).
The occurrence of each taxon was mapped on the ETP.When there was a lack of geographic information of the species record or the accuracy of the data was larger than 1° of latitude by 1° longitude, the information was discarded.Once the distribution of each species was mapped on the ETP, we traced a grid of 1° of latitude by 1° longitude and obtained the species richness per grid.Since not all records have accurate geographic data, the amount of information included in the following analysis is slightly reduced (~0.03%)than the published information.Nonetheless, mapping of species distribution was a methodological prerequisite to generate an accurate picture of the knowledge and gaps of the marine biodiversity of the benthic caridean shrimp species in the ETP.In addition, occurrence of spatially rare (inhabiting 1 to 2 grids) species in the area was also mapped.After mapping, we fitted species richness and species richness of rare species (inhabiting 1 to 2 grids) against latitude through a linear model.Additionally, we related the number of caridean species against the number of caridean references (proxy for sampling intensity) per grid with a Spearman Rank correlation.
We used the species occurrence per grid to generate sample-based rarefaction curves of observed species and calculated the species richness estimator Chao2 for the ETP (Colwell, 2009).While the former analysis indicates the observed species accumulation curve, the latter is a sample-based nonparametric estimator for addressing expected species richness of the benthic caridean shrimp in the Eastern Tropical Pacific.EstimateS (Colwell, 2009) were used to compute observed and expected rarefaction curves and their confidence intervals (95%) after 10,000 randomizations with replacement.

resulTs
The present checklist of the marine benthic caridean shrimp of the Eastern Tropical Pacific includes 183 species considered as currently valid, belonging to 67 genera, 18 families and 9 superfamilies (Tab.1).
Historic accumulation curves varied along the ETP, nevertheless, some results were similar among countries (Fig. 2).For example, for countries such as Mexico (Fig. 2A), Ecuador (Fig. 2B) and Panama (Fig. 2C), small but steady early increases were followed by sudden increments in species richness.Likewise, the same trend is recorded when data are pooled across the ETP (Fig. 2H).The observed trend was meaningfully fitted with an exponential function, indicating that during the last three decades many more species records were added to the benthic caridean fauna than those recorded in the ETP during the previous 110 years.Function fitting also indicates that the current increment of species records still not leveling off for any particular country neither for the entire ETP (Fig. 2).
According to species mapping, richness varies largely among grids in the ETP (Fig. 3).Except for a few empty spots in the northern Gulf of California, the Tropical Mexican Pacific, Guatemala and Honduras, marine benthic caridean shrimp have been recorded in the coastal margin of the entire ETP and its oceanic islands (i.e., Revillagigedo, Clipperton, Malpelo, Cocos,

Caridean shrimp from the Eastern Pacific
Nauplius, 26: e2018035 Galapagos).Species richness is concentrated in grids located around Galapagos and Panama (i.e., Central America), but relatively rich spots can also be observed in Colombia, Costa Rica and the Gulf of California, hence disrupting any meaningful trend between richness and latitude (r 2 = 0.002, n = 37, p> 0.05).
Data indicates that up to 44 species (24%) inhabit a single grid and hence may be considered spatially rare, but as many as 135 species (~74%) are restricted to less than five grids each.In contrast, just 22 species (12%) inhabit more than 10 grids and hence may be considered as spatially common.The species Processa peruviana Wicksten, 1983 (15 grids) Rare species, inhabiting one to two grids, concentrated around Central America (Costa Rica, Panama, Colombia and Galapagos Islands) and the entrance of the Gulf of California, Mexico; however, from northern Costa Rica (~11°N) up to the Central Mexican Pacific (~19°N) spatially rare species are lacking (Fig. 4).Overall, linear regression models indicates no latitudinal trend (r 2 = 0.001-0.004) in the number of rare species (i.e., inhabiting 1 to 2 grids).
Finally, the non-parametric Chao2 species estimator indicates that the observed species richness is lower than expected (Fig. 5), in fact with a 95% confidence the model suggests that the expected species richness of marine benthic caridean shrimp in the ETP may lay somewhere between 202-269 species, a number 11%-47% larger than the observed.
Data revealed that families Alpheidae and Palaemonidae, and genus Alpheus contributed to most of the species.Dominance of the former two families was previously recognized by Wicksten and Hendrickx (1992) and Anker et al. (2006b).The trend, however, is not exclusive to the ETP but it is also observed in the Caribbean and the Indo West Pacific (Boschi, 2000).Bauer (2004) and Poupin (2008) have suggested that Alpheidae and Palaemonidae dominance is related to its ability to inhabit a large number of substrates, its reduced size and the capacity of its individuals to establish symbiotic relations with vertebrates or other invertebrates.In other invertebrate and vertebrate taxa such as ophiuroids (Granja-Fernández et al., 2014) and gobies (Zepeta-Vilchis et al., 2013), such attributes have also been signaled as responsible for increasing species richness and dominance of the groups in marine habitats.On the contrary, taxa included on Pasiphaeidae or Pandalidae members of which often constitute an important component of the mesopelagic and bathypelagic planktonic communities (Komai et al., 2012) are relatively larger than Alpheidae and Palaemonidae, and does not establish symbiotic relations with other taxa and hence not as species rich as Alpheidae.
Grid data indicates that species richness concentrates in Galapagos and Panama, but relatively rich spots can also be observed in Colombia, Costa Rica, and the Gulf of California.Boschi (2000) and Wicksten and Hendrickx (2003) recognized the Panamic Province as the most diverse caridean shrimp area in the entire Eastern Tropical Pacific.However, if we further analyze the number of caridean species against the number of caridean references per grid, we will notice that rich spots correspond with the most heavily studied areas in the ETP; in fact there is a meaningful positive relationship (Spearman rank correlation r = 0.88, p = 0.00) between species richness and the number of caridean references per grid.While the number of references is not an adequate proxy for sampling intensity and a correlation is not a cause-consequence proof, the observed relation suggests to some extent that the observed benthic caridean species richness in the ETP may have not exclusively resulted from habitat heterogeneity or large-scale evolutionary processes currently invoked as one of the several speciation or richness drivers in the marine realm (Tittensor et al., 2010), instead they may have also arisen from differences in research effort and hence from sampling bias.In the same way, spatial analyses of rare species are not meaningfully related to latitude in the studied area (i.e., no meaningful increase toward Central America).The observed results, rather than plea against the currently recognized caridean spatial richness pattern (Boschi, 2000;Wicksten and Hendrickx, 2003), call for a more thoughtful and systematic sampling protocol in order to get information from poorly sampled areas in the Eastern Tropical Pacific.
The function that explains the historic patterns per country and for the entire region suggest that the inventory of the marine benthic caridean shrimp fauna from the ETP is still incomplete.The extent of the gap between the observed and the expected caridean diversity in the ETP is unknown, but several evidence lines may suggest its size.First, if we considered that the benthic caridean records are restricted to an area that occupies 8.8% of the entire ETP, it becomes obvious that most of the region (~91%) is still not inventoried.In this regard if the species-area relationship is invoked, then marine benthic caridean shrimp fauna of the ETP must be nine times larger than the observed and hence is still far from being complete.According to

Caridean shrimp from the Eastern Pacific
Nauplius, 26: e2018035 the species-area relationship and the Chao2 species estimator, the marine benthic caridean shrimp fauna of the ETP must be 11% to 900% larger than the current observed.Species-area relationships have been invoked to explain species richness for several biological groups (Connor and McCoy, 2001) including marine invertebrates (Alvarado et al., 2010;Costello et al., 2010;Granja-Fernández and López-Pérez, 2012).Nonetheless, this number must be considered with caution since habitat diversity, area per se, passive sampling and edge effects are not mutually exclusive mechanisms and may operate individually and/ or in synergy to cause and or modify species-area relationships (Connor and McCoy, 2001).On the other hand, the non-parametric Chao2 species index indicates that caridean richness is 11% to 47% larger than recorded; this number, however, is sensibly minor compared to the suggested by species-area relationship, though it is in agreement with estimations regarding the entire crustacean fauna from the Mexican Tropical Pacific using different approaches (González-Oreja et al., 2010;García-Madrigal et al., 2012;López-Pérez et al., 2012).In the absence of complete inventories, nonparametric estimators (for example, Chao2) have been shown to perform better than most other methods, such as observed species richness, speciesarea curves or asymptotic estimators (Colwell, 2009), but the Chao2 estimator still gives a lower bound to expected species richness, thus producing conservative estimates (Hortal et al., 2006).
It is revealing that 44 species (24%) inhabit a single grid (as much as ~11,200 km 2 ) but as many as 135 species (~74%) are restricted to less than five grids each.This either indicates that most of the caridean shrimp species in the ETP have a narrow distribution (i.e., provincialism) or that current distribution of narrowly distributed species (74%) resulted from sampling deviations.Previous statements about endemism or narrowness in species distribution in taxonomic groups such as corals, crustaceans, echinoderms or fishes in the ETP indicating that species are spatially restricted to Central America or the Gulf of California, have been contradicted by recent genomic analyses, detailed taxonomic studies or intensive sampling protocols across the area (Combosch et al., 2008;Zepeta-Vilchis et al., 2012;Granja-Fernández et al., 2013;Paz-García et al., 2015;Reyes-Bonilla et al., 2017;Valencia-Méndez et al., 2017).While work on this regard is pending on ETP caridean fauna, the relatively large sampling gaps (~ 91%, see above) regarding the group is the most likely scenario for explaining the narrow distribution of a large number of species in the area, not only for caridean shrimps (this contribution) but across several taxonomic marine groups and areas worldwide (Costello et al., 2010).
Finally, integrative databases (such as the current one) resulting from previously published information to make generalizations in ecology, without any doubt is subjected to errors that distort our view.In particular, bias in sample size, but also in space, time and taxonomy may have altered the stated generalizations about the species richness distribution pattern on ETP caridean fauna; nevertheless the current metaanalysis has an advantage over simple verbal reviews because unbiased estimates of the overall strength of a hypothetical relationship can be obtained, heterogeneity in research findings identified and new research questions developed (Møller and Jennions, 2001).In this regard, detailed taxonomy of the Caridea and intensive sampling across environments in the ETP still is the most urgent need in the studied area.

Figure 3 .
Figure 3. Species richness of marine benthic caridean shrimps per grid (1° latitude by 1° longitude) in the Eastern Tropical Pacific.Notice color scale.

Figure 5 .
Figure 5. Species richness of marine benthic caridean shrimps in Eastern Tropical Pacific.Orange continuous line = observed richness, blue continuous line = expected richness, blue long-cut lines = confidence interval (95%).Expected species number and confidence intervals (95%) were generated after 10,000 randomizations with replacement using non-parametric Chao2 estimator.

aCKnoWledgeMenTs
Authors are in debt to Michel Hendrickx and Mary Wicksten for providing critical references, and to Arthur Anker, Sammy De Grave, Michel Hendrickx, Mary Wicksten and Carolina Tavares for taxonomic advice and critical review of the checklist.Authors are grateful to Socorro García (UMAR) and Luis Hernández (UABCS) for early comments to the manuscript and to anonymous reviewers for suggestions and corrections during revision process.While writing this manuscript BMG received a graduate fellowship (257618) from Consejo Nacional de Ciencia y Tecnología (CONACYT), Mexico.