A First Estimate of Species Diversity for Benthic Diatom Assemblages from the Revillagigedo Archipelago, Mexico

Recent investigations at previously unexplored localities on the Mexican coast have confirmed the high taxonomic potential of benthic marine diatom assemblages (BMDA) in the region. An exploratory study of epiphytic diatoms of macroalgae in the Revillagigedo Archipelago (RA) suggested that further studies would yield many more taxa, prompting the hypothesis that diversity measurements, based on ecological indices, would be among the highest in pristine environments. Thus, the aim of this research was to enrich the record of epiphytic diatom floristics of the RA, and to estimate species diversity based on information theory (H′). Floristically, 167 identified taxa are added here to the BMDA species list of the RA, bringing the total to 397 taxa overall, including 52 taxa that are potentially new records for the Mexican Pacific coast. Among the most conspicuous genera are Mastogloia with five new taxa and it remains the most diverse genus with 55 taxa overall, followed by Cocconeis (27), Nitzschia (24), Amphora (23), Navicula (19), Diploneis (17) and Grammatophora (15). As expected for a pristine environment, the computed species diversity values for the BMDA were high, ranging from H′ = 3.92–5.2, depicting stability. Future surveys are expected to further increase the species richness of BMDA for the RA.


Introduction
The ecological significance of benthic diatoms has distinguished them as a suitable referential for estimating biodiversity in marine protected areas [1] and as useful references for assessing environmental impact [2]. Thus, both diversity and ecological aspects of benthic diatoms should be considered essential in order to have an adequate ecological prospect of any aquatic ecosystem, particularly when developing management plans for protected areas. The high floristic potential of benthic diatoms (Bacillariophyceae) for Mexican littorals has recently been confirmed through surveying new substrata and by studies in hitherto unexplored areas that have yielded numerous new records for the region [3][4][5][6][7] and have enriched the overall species list for the region [6]. According to [8], in some parts of the biosphere where diatoms are abundant, there is very little known about them. Thus, a good case can be made that at least some diatom species and even a few genera are endemics, but many such claims are still weak. Such claims about diatom distributions and endemism must be treated with skepticism.
In this sense, the potential endemism and particularities of benthic diatom taxocenoses from remote locations such as the Revillagigedo Archipelago (RA) encouraged the first floristic study of epiphytic diatom assemblages on macroalgae from that region, yielding a species richness of 208 diatom taxa [6].

Materials and Methods
The Revillagigedo Archipelago comprises San Benedicto, Socorro, Roca Partida and the Clarion islands located 386 km south of Cabo San Lucas, Baja California Sur, under the jurisdiction of the state of Colima, México. In July 2016, UNESCO declared it a World Heritage Site [13], and recently there was an ex professo study for justifying its declaration as a national park [14,15]. Seasonal variations in the RA region seem to be determined by the alternating influence of the California Current and the North Equatorial Current, whilst the rest of the year a transition between both states may be observed [16]. The archipelago rocky coasts harbor a rich macroalgal community of 190 to 200 macroalgae taxa [11,17], and on the surfaces of these macroalgae, floristic rich assemblages of epiphytic diatoms have been identified [6]. Environmental conditions suggest that the biogeographical affinity of the epiphytic diatom taxocenoses is strongly tropical, although has important temperate components [6].
During an opportune (recreational) trip to the Revillagigedo Islands, from 10 to 15 February 2019, 15 macroalgae thalli were collected manually by scuba diving at 7-27 m depths on the rocky shore of San Benedicto and the Socorro islands (RA).
Specific sampling points included El Boiler ( and identified in the laboratory following [18]. Diatoms were brushed off from each algae specimen while rinsing with purified water, washing the brush after each use. The brushed-off material was placed in a 150 mL test tube and left to settle. Thereafter, the precipitates were collected and oxidized using a mixture of commercial alcohol and nitric acid at a ratio of 1 (sample): 2 (alcohol): 5 (acid), according to [2]. Thereafter, the oxidized material was rinsed repeatedly with purified water until it reached a pH ≥ 6. After that, for each sample, two double permanent slides were mounted using the synthetic resin Pleurax (RI = 1.7). The slides were inspected under several magnifications (including 250×, 400×, 630× and 1000×) for general recognizance of diatom taxa and forms. Diatom identification was conducted at 1000× under an Olympus CH-2 compound microscope with brightfield illumination, following [1, 2,6,. Taxonomic status was updated according to the Algaebase Web site at http://algaebase.org/search/species/ (accessed on 1 August 2021) [46]. Images of selected specimens were captured with a CMOS Konus digital lens at 630× (oil) and 1000× (oil) and used to construct an iconographic catalog to back the reliability of the identifications. Thus, images of previously recorded taxa in [6] were also included that show no correspondence with the list of new additions to the Revillagigedo diatomological flora but were observed during the quantitative analysis.
Specific sampling points included El Boiler (19°19′38.30″ N, 110°48′17.16″ W) and El Cañon (19°17′40.48″ N, 110°48′40.92″ W) on San Benedicto, and Cabo Pearce (18°45′57.24″ N, 110°54′00.30″ W) and Punta Tosca (18°46′33.46″ N, 111°03′26.01″ W) on Socorro Island (Figure 1). The macroalgae specimens were sun dried, transported in plastic Ziploc bags, and identified in the laboratory following [18]. Diatoms were brushed off from each algae specimen while rinsing with purified water, washing the brush after each use. The brushed-off material was placed in a 150 ml test tube and left to settle. Thereafter, the precipitates were collected and oxidized using a mixture of commercial alcohol and nitric acid at a ratio of 1 (sample): 2 (alcohol): 5 (acid), according to [2]. Thereafter, the oxidized material was rinsed repeatedly with purified water until it reached a pH ≥ 6. After that, for each sample, two double permanent slides were mounted using the synthetic resin Pleurax (RI = 1.7). The slides were inspected under several magnifications (including 250×, 400×, 630× and 1000×) for general recognizance of diatom taxa and forms. Diatom identification was conducted at 1000× under an Olympus CH-2 compound microscope with brightfield illumination, following [1, 2,6,. Taxonomic status was updated according to the Algaebase Web site at http://algaebase.org/search/species/ (accessed on 1 August 2021) [46]. Images of selected specimens were captured with a CMOS Konus digital lens at 630× (oil) and 1000× (oil) and used to construct an iconographic catalog to back the reliability of the identifications. Thus, images of previously recorded taxa in [6] were also included that show no correspondence with the list of new additions to the Revillagigedo diatomological flora but were observed during the quantitative analysis.  Species diversity estimations of the epiphytic diatom assemblages were based on information theory [47,48]. Values of Shannon's index H using log 2 in the form of bits/taxon were computed, supported by species richness and values of other diversity indices such as Pielou's equitability (J ), Simpson's dominance (λ) and diversity (1 − λ) [12]. For this, based on diatom abundances observed in the mounted slides, the samples from 12 sites were selected, and relative abundances of all diatom taxa estimated based on a sample size (N) of 500 valves per mounted slide (sub-sample), including a replicate. Finally, the similarity between all the samples and their replicates was measured using the Bray-Curtis indices to gain an insight into the taxa distribution among the sampling sites relying first on solely the presence/absence of taxa and considering their relative abundances. Clusters were derived from the Bray-Curtis similarity values from each matrix. All calculations were performed using program Primer 6 [49].
During the quantitative analysis, it was noted that the abundance of valves in all the mounted preparations was low. Notwithstanding, 12,443 valves were counted, and 276 taxa were included, 47 of which were Mastogloia species. Eighty percent of the counted valves belonged to only 35 taxa (Table 2), while 20% of the valves comprised the other 241 taxa. The most frequent were Achnanthidium exiguum var. heterovalvata, Cocconeis krammeri, Mastogloia crucicula, Cocconeis scutellum var. parva and Epithemia pacifica ( Table 2). The number of taxa per sub-sample (replicate) ranged from 47 to 87 taxa, with a maximum spread of 16 between sub-samples from the same locality. According to their relative abundances, either as independent taxocenoses of each sampling locality or an overall taxocenoses, the diatom assemblages showed a typical structure of a few abundant taxa, more common taxa and many uncommon and rare taxa. Said classification is arbitrary in terms of the number of valves for each category but fits the whole species list with most of the taxa being uncommon and rare (88%), and thus excluded from Table 2.                                       Notwithstanding the depicted distribution, the average species diversity value was H = 4.68 bits/taxon, ranging from high H = 3.92 (S = 55) to very high 5.2 (S = 87). Furthermore, between the lowest and richest sub-samples (replicates) the H values were both high, 4.11 (S = 57) vs. 4.62 (S = 73), with all community parameter values showing good correspondence between them, particularly with low values of dominance, high values of equitability and species richness (Table 3). In terms of similarity, measurements for all the collected samples indicate a dissimilarity between the different diatom assemblages, showing around or over 50% similarity based on the presence/absence of taxa; this is compared to the values between replicates that depict the same assemblage and varied at around 0.7 similarity ( Figure 36). Whilst, based on the taxa relative abundances, values ranged from just below to just above 0.3 ( Figure 37), this is compared to the similarity between replicates which varied at around 0.8. All this indicates that much of the taxa, which compose the assemblages, are different including the most numerous ones.

Discussion
With the new records from this study (167), the species richness of diatoms found on macroalgae from the Revillagigedo Archipelago reaches a total of 397 taxa. This implies an increase of >75% in this report alone, not including the independently recorded species of the genus Mastogloia whose species richness had previously increased by approximately 75% [10]. Thus, an overall increase in species richness of 90%, including the previously recorded Mastogloia taxa, is attained.

Discussion
With the new records from this study (167), the species richness of diatoms found on macroalgae from the Revillagigedo Archipelago reaches a total of 397 taxa. This implies an increase of >75% in this report alone, not including the independently recorded species of the genus Mastogloia whose species richness had previously increased by approximately 75% [10]. Thus, an overall increase in species richness of 90%, including the previously recorded Mastogloia taxa, is attained.
Considering the previous study of the RA [6], the five new Mastogloia taxa recorded make this genus the most diverse, with 55 taxa in the study area, followed by the Cocconeis (27), Nitzschia (24), Amphora (23), Navicula (18), Diploneis (17) and Grammatophora (15). It does seem valid to remark that such an increase in species richness, which is only four less than the predicted number, supports our hypothesis on the potential species richness of benthic diatoms in the RA. Moreover, some of the taxa were also recorded recently from the Gulf of California, such as Achnanthes apiculata, Caloneis egena, Coronia ambigua, Grammatophora undulata var. gallopagensis, Gyrosigma parvulum, Psammodictyon pustulatum, Psammodictyon roridum, Seminavis basilica, Trachyneis aspera var. oblonga and Tryblionella nicobarica, among others [7,50]. This indicates a biogeographical connectivity between regions where warm conditions determine the presence of species with a tropical affinity. It thus seems that longdistance dispersal is, in many cases, effective enough for species to establish across large geographic areas [8]. This, however, should also be examined relying on the Mastogloia taxa that they share.
Regarding other taxonomic issues, in certain cases apparent synonymies in the identified taxa were not updated and the names from the literature were kept. These included Amphora novaecaledonica, which is deemed a synonymy of Amphora ostrearia var. vitrea Cleve [46]. Our specimen, however, closely resembles the A. novaecaledonica in [39] but differs from A. ostrearia var. vitrea in several modern references. Additionally, our Nitzschia punctata var. coarctata is considered synonymous with Tryblionella coarctata (Grunow) D.G. Mann, which was recorded in the previous floristics for the RA [6]; the forms were different, and the assigned names were consistent with several modern references. Furthermore, in the case of Thalassionema nitzschioides var. parvus, we opted for this authority vs. Heiden and Kolbe 1928 [46], provisionally.
On the other hand, also backing our hypothesis are the species diversity values which ranged from H = 3.92-5.2 bit/taxon, with an average value of 4.68. Typical diatom assemblages are shaped by a few abundant and very common species, and many rare and uncommon species. Variations in this distribution pattern are reflected in the mathematical values of diversity derived from information theory calculated with Shannon's index (H ). Estimated values of H have been observed to vary mainly between (modal) values of 2.6-3.8 bits/taxon (median, 2.4-4.6) benthic diatom assemblages [51] and have been interpreted as usual or moderately high values of diversity that indicate stability of the assemblages. Although higher (H ≈ 5) and lower (H < 2) values are not uncommon, these have been interpreted as indicative of a tendency towards an improbable and thus unstable state of the assemblages in nature [12]. Thus, the primary high estimates of species diversity for the inspected benthic (epiphytic) diatom assemblages of the RA may reflect the pristine conditions of the islands due to the combined factors of remoteness and enforced policies as a protected natural area. When compared to diversity values from other protected areas, such as the Guerrero Negro lagoon, certain equivalence is noted with the average values of H = 4.96 bits/taxon, and range of 3.7-5.9, computed for the epipelic diatom assemblages of that lagoon [3]. So, according to the latter study, the high values of diversity depict pristine or undisturbed conditions (independent of whether the habitat is protected or not), but species richness and the corresponding diversity values are expected to be as high as those in both studies. Whatever the case, the assessed attributes of the diatom assemblages inspected indicate that the Revillagigedo Archipelago is free from any environmental impact.
Finally, the low similarity between various compared samples is typical of benthic diatom taxocenosis that exhibit a patchy or aggregate distribution in which the species composition and the main taxa that they are composed of are both different and are most likely alternating these components on a successional basis [2]. More precise descriptions of space distribution require a hypothesis-based sampling design, preferably coupled with temporal variations of the diatom assemblages.
In terms of limitations of this study, the 29 taxa that could not be identified are added to those in the previous survey [4], making a total of 61 taxa for the RA that are expected to be undetermined species requiring formal taxonomic treatment. Some of the reference images are deemed in need to of improving and further sampling is likely to provide better specimens to replace them. Meanwhile, they are considered to adequately serve as backup for the identifications made.
On the other hand, the generated ecological parameter estimates are based on the inspection of a few host specimens with very few macroalgal taxa (<10) out of the >200 macroalgal taxa recorded from the RA [17]. Besides this, other substrates such as sediments or rock or a number of other localities in the area were not considered. Thus, the potential number of taxa [6] and concomitant ecological diversity parameter values that can be expected by inspecting a more representative number of samples from the aforementioned substrates may surpass the S and H values observed in this study.
Furthermore, the taxonomic relevance on classification, determination and identification issues of benthic diatoms in these types of environments show the benefits of focusing on remote protected areas worldwide. The bearing that this observation can have on ecological and biogeographical topics clearly justifies further research of benthic diatoms in these protected areas.