Introduction

Macroalgae are an ideal substratum for many species of Bacillariophyceae. Over two hundred diatom taxa living as epiphytes on multispecies macroalgal assemblages have been recorded for the southern Baja California peninsula, both along the west coast (^{Siqueiros Beltrones and Valenzuela Romero, 2004}; ^{Siqueiros Beltrones and López-Fuerte, 2006}) and the east coast (^{Siqueiros Beltrones and Hernández Almeida, 2006}). However, a high number of diatom taxa has also been recorded living on a single species of macroalgae, such as in the giant kelp Macrocystis pyrifera L. (C. Agardh) where the number of diatom species surpassed 170 taxa for several sampling dates (^{Siqueiros Beltrones et al., 2002}; ^{Siqueiros Beltrones and Argumedo-Hernández, 2005}; ^{Argumedo-Hernández and Siqueiros Beltrones, 2008}). Moreover, in a recent study where a single specimen of the rhodophyte Ploclamiun cartilagineum (Lamoroux) Dixon was examined, as many as 42 diatom taxa were recorded (^{Siqueiros Beltrones and Argumedo-Hernández, 2014a}).

Multiple species of diatoms colonize macroalgal substrates and are ingested along with their hosts by many species of grazers, including abalone (Haliotis spp.). In particular, the main interest surrounding epiphytic diatoms of M. pyrifera comes from its role as the main food-source for several abalone species whose diets depend strongly on its heavily epiphytized blades, which is the most abundant kelp on the western coast of the Baja California peninsula (Mexico).

Based on the above, we were interested in determining which diatom taxa were to be found living on other macroalgae that are also grazed by abalone (^{Siqueiros Beltrones et al., 2002}); the phaeophyte Eisenia arborea Areschoug seemed as the next-best option, inasmuch as it is the second most abundant kelp along the western coast of the peninsula (^{Hernández-Carmona et al., 2009}), and would represent an ecological alternative, particularly when the M. pyrifera forests disappear during El Niño events. Like M. pyrifera, E. arborea is distributed from British Columbia, Canada, down to Bahía Magdalena, Mexico, in the intertidal zone where it forms dense beds over rocky substratum (^{Pedroche et al., 2008}). We thus assumed that the epiphytic diatom assemblage it harbored would be similar to that of M. pyrifera in species composition and structure, i.e., species richness, diversity, dominance. However, the first ex profeso observations made on E. arborea blades from the same region did not support such expectations because diatoms were not found (^{Siqueiros Beltrones et al., 2002}). Nonetheless, the presence of diatoms on blades of E. arborea from the same region was recently noted when samples of this kelp were processed for biochemical assays (Muñoz, M., IPN-CICIMAR, com. pers.). After verifying the report, we established our objective to describe the diatom assemblages on the basis of their species composition and association structure. Thus, based on previous observations in epiphytic diatoms of M. pyryfera (^{Argumedo Hernández and Siqueiros Beltrones, 2008},) we tested the hypothesis that, on the blades of E. arborea from different dates, diatom taxa representing distinct stages of succession would be found together with either abundant pioneer taxa such as Cocconeis spp., or naviculoid opportunistic species that depict advanced stages of colonization, in both cases associated with low values of diversity. We conclude that the structure of the diatom assemblage is unusual, similar to those of assemblages from extreme environments. However, structure analysis did not permit the distinction of succession stages in the epiphytic diatom assemblage.

Materials and Methods

Blades of E. arborea were collected at Punta Arenas (24º 33´ 2´´ N and 112º 05´ 28´´ W), off Isla Magdalena (Fig. 1), in the west coast of B.C.S. Samplings were gathered in September, November, and December, 2013, and in January, March, and May, 2014, at a depth of 5-10 m by scuba diving. The blades were dried before being transported to the laboratory, where the most noticeable epiphytized blades were selected among those having few or no bryozoans. Epiphytes on the surface of the E. arborea blades from each date were scraped off using a glass slide while rinsing with purified water into a dissecting tray. The resulting epiphyte concentrate was stored in assay tubes; decanted water was eliminated and then a sample of each concentrate was oxidized in assay tubes using a mixture of commercial ethanol and nitric acid at a ratio of 1:3:5 that varied depending on the amount of organic matter in each sample (^{Siqueiros-Beltrones, 2002}). Afterwards, repeated rinsing was done using purified water until a pH > 6 was reached. From each cleaned diatom sample five (double) permanent preparations and a repetition were mounted using the synthetic resin Pleurax (RI=1.7).

Figure 1 Location of sampling site Punta Arenas off Isla Magdalena, B.C.S., México. 

Diatom taxa were identified under a phase contrast compound microscope with planapo-chromatic lenses at 1000×, following the references of ^{Cleve-Euler (1968)}; ^{Desikachary (1988)}; ^{Hendey (1964)}; ^{Hustedt (1959}; ^1961-66); ^{López-Fuerte et al. (2010)}; ^{Peragallo and Peragallo (1908)}; ^{Round et al. (1990)}; ^{Schmidt et al. (1874-1959)}; ^{Siqueiros Beltrones (2002)}; ^{Siqueiros Beltrones and Valenzuela-Romero (2001)}; ^{Siqueiros Beltrones et al. (2004)}; ^{Siqueiros Beltrones and Hernández-Almeida (2006)}; ^{Witkowski et al. (2000)}. A representative micro-photographic record of the diatom taxa is provided.

In order to describe the structure of the epiphytic diatom assemblages of E. arborea, the relative abundances of the taxa were estimated based on an approximate sample size (n) of 500 valves (^{Siqueiros Beltrones, 2002}) per sample (N = 3600). With these, the following indices for estimating diversity were calculated: Shannon´s (H´), Pielou´s (J´), Simpson´s (l); and the Bray-Curtis index for measuring similarity between samples, which was complemented using the Jaccard index. All computations were done using program Primer 6 v 6.1.6.

Results

Floristics. The resulting floristic list shows a total of 99 diatom taxa living on blades of E. arborea (Table 1). A photographic catalogue including most of the taxa is provided (Figs. 4-158). Most taxa, either common or rare, are epiphytic forms, except for certain taxa like the Lyrella forms and Diploneis spp. that are epipelic. Others, such as the centric forms Biddulphia biddulphiana (Boyer) J.E. Smith, Paralia sulcata f. radiata Grunow, Paralia sulcata var. crenulata Grunow, and Podosira stelliger (Bailey) Mann are typical epiphytes of the region; biraphids forms such as Campylopyxis garkeana (Grunow) Medlin, Gomphonemopsis pseudexigua (Simonsen) Medlin, Gomphoseptatum aestuarii (Cleve) Medlin, and Rhoicosphenia genuflexa (Kützing) Medlin are characteristic epiphytes of M. pyrifera and were proportionally represented; however, others such as Grammatophora spp., which are common epiphytes were not as conspicuous as in other macroalgae of the region. On the other hand, the araphid Pteroncola inane (Giffen) Round (^{Siqueiros Beltrones and Argumedo Hernández, 2014b}) and the centric Biddulphia grundleri A. Schmidt constitute new records for the Mexican NW region. The former was extremely abundant in most samples, while the latter was rare.

Table 1 Diatom taxa living on blades of Eisenia arborea off Isla Magdalena, B.C.S., Mexico collected from September 2013 to May 2014. * New record for the Mexican NW region; **Record reported in ^{Siqueiros Beltrones and Argumedo Hernández (2014b)} for the same survey. 

Figure 2 Similarity between samples of epiphytic diatom assemblages living on blades of Eisenia arborea from B.C.S., measured with the Jaccard index. 

Figure 3 Similarity between samples of epiphytic diatom assemblages living on blades of Eisenia arborea in B.C.S., measured with the Bray-Curtis index. 

Figures 4-5 4) Biddulphia biddulphiana; 5) Biddulphia grundleri. Scale bar =10 μm for both specimens. 

Figures 6-14 6, 9) Coscinodiscus radiatus; 7) Coscinodiscus rothii; 8) Azpeitia nodulifer; 10) Coscinodiscus crenulatus; 11-12) Roperia tesselata; 13) Thalassiosiraeccentrica; 14) Campylodiscus cf. simulans. Scale bar = 10 μm for all specimens. 

Figures 15-25 15) Paralia sulcata var. crenulata; 16-17) Paralia sulcata f. radiata; 18-20) Cyclotella litoralis; 21) Licmophora gracilis; 22, 25) Psammodiscus nitidus; 23) Actinoptychus aster; 24) Actinoptychus senarius. Scale bar = 10 μm for all specimens. 

Figures 26-33 26, 28-30) Actinocyclus curvatulus; 27) Campyloneis grevillei; 31) Raphoneis surirella var. ceylanica; 32) Delphineis surirella var. australis; 33) Asteromphalusarachne. Scale bar = 10 μm for all specimens. 

Figures 34-56 34-35) Grammatophora macilenta; 36-37, 48) Grammatophora oceanica; 38-39, 41, 44-46, 56) Grammatophora marina; 40, 42-43) Grammatophora hamulifera; 47) Grammatophora marina var. subundulata; 49) Grammatophora undulata; 50, 52-55) Pteroncola sp.; 51) Synedra barbatula. Scale bar = 10 μm for all specimens. 

Figures 57-85 57) Cocconeis vetusta; 58, 59-62, 75) Cocconeis dirupta var. flexella; 59) Cocconeis contermina; 60) Cocconeis krameri; 63-64) Cocconeis scutellum; 65, 74) Amphicocconeis disculoides; 67) Cocconeis convexa; 68-70) Cocconeis pseudodiruptoides; 69) Cocconeis diminuta; 66, 72-73) Cocconeis speciosa; 76) Diploneis aestuari; 77) Cocconeis latecostata; 78) Achnanthes manifera; 79-80) Berkeleya rutilans; 81-83) Rhoicosphenia genuflexa; 84) Achnanthes yaquinensis; 85) Mastogloia macdonaldii. Scale bar = 10 μm for all specimens. 

Figures 86-100 86) Diploneis crabro; 87-88) Diploneis bombus; 89-90) Diploneis papula; 91) Diploneis obliqua; 92) Diploneis litoralis; 93) Diploneis papula var. constricta; 94) Lyrella exsul; 95) Rhoicosphenia genuflexa; 96-97) Diploneis didyma; 98) Diploneis smithii; 99) Berkeleya rutilans; 100) Navicula clavata var. caribaea. Scale bar = 10μm for all specimens. 

Figures 101-119 101, 103) Trachyneis velata; 102, 104) Trachyneis aspera; 105) Campylopyxis garkeana; 106) Navicula cf. mollissima; 107) Navicula incerta; 108 Staurophora sp.; 109) Navicula cf. feuenbornii; 110, 114) Navicula cincta; 111) Navicula cancellata; 112) Navicula pennata; 113) Navicula directa; 115) Navicula diversistriata; 116) Navicula sp. 1; 117) Parlibellus sp.; 118-119) Campylopyxis garkeana. Scale bar = 10 μm for all specimens. 

Figures 120-142 120) Seminavis ventricosa; 121) Amphora angusta; 122, 130, 135, 137) Amphora proteus var. contigua; 123) Amphora angustissima; 124, 126, 129) Campylopyxis garkeana; 125, 127-128, 134) Caloneis linearis; 131) Amphora holsaticoides; 132-133, 141-142) Gomphoseptatum aestuarii; 138-140) Gomphonemopsis pseudexigua. Scale bar = 10μm for all specimens. 

Figures 143-158 143-146) Nitzschia sicula; 147, 152) Nitzschia distans; 148) Tabularia tabulata var. fasciculata; 149) Hyalosynedra laevigata; 150, 158) Tabularia investiens; 151) Nitzschia hybrida; 153) Nitzschia sp. 1; 154-155) Denticula kuetzingii; 156) Nitzschia punctata var. coarctata; 157) Anaulus cf. minutus. Scale bar = 10 μm for all specimens. 

As implied above, the epiphytic diatom assemblage on E. arborea was characterized by the extreme dominance by a single taxon, P. inane (Figs. 50, 52-55), which represented between 33% and 94% of the counted valves in the different dates, and more than 66% overall (N = 3600); by contrast, the second abundant taxon, Campylopyxis garkeana (Figs. 105, 118, 119) represented less than 5% of the total valves (Table 2). Moreover, the quantitative phase yielded only 54 species, most of the taxa being rare or uncommon (Table 1). In spite of the above, according to the estimated relative abundances, the described diatom assemblage exhibits a typical pattern, i.e., many rare and uncommon species, and few abundant (or very common) ones. In this case, however, it is unusual that a single species contributes 66% of the total abundance.

Table 2 Relative abundance of the main diatom taxa in the epiphytic assemblages living on blades of Eisenia arborea collected in B.C.S., Mexico from September 2013 to May 2014. 

Species diversity. The estimated values of species diversity ranged in general from low to very low (Table 3), with the lowest value corresponding to the September sample (H´ = 0.3; S = 6), in which Pteroncola inane was more abundant. However, in other samples where the cell concentration in the preparations was either abundant (January) or very low (April, May), due to P. inane, diversity values were also very low; unlike the diversity values of the samples from November, December, and March (Table 3), which are similar to those from normal (favorable) environments, as a result of their species richness (S = 24, 28, 24, respectively) and equitability (J´) higher than 0.5, which are in fact atypically lower.

Table 3 Diversity values estimated for the epiphytic diatom assemblages found on blades of Eisenia arborea in Isla Magdalena, B.C.S. from September 2013 to May 2014. S= species richness; N= valves counted; J´= equitability; H´= Shannon´s diversity; l= Simpson´s dominance; 1- l=Simpson´s diversity. 

Similarity. The similarity values between samples of epiphytic diatom assemblages living on blades of Eisenia arborea based on presence/absence of species (Jaccard´s index) show that September is the most distinct sample (<20%), mainly because of the absence of taxa (S = 6). The rest show a similarity of around 40% which, according to other studies and the characteristics of the index, is only slightly low (Fig. 2). In contrast, when also considering the relative abundance of each taxon (Bray-Curtis index), the minimum similarity was 60% (Fig. 3), but was generally high (approximate 80%), mainly due to the abundance of P. inane in most of the samples; the segregation of the December sample, however, has to do with an increase in the abundance of C. garkeana (Table 2).

The main components in the low values of diversity were thus the high abundance of P. inane, the scarcity of the other taxa (low S), and the low equitability. These parameters did vary between the samples (dates) but were in general atypical of benthic diatom assemblages; such low values resemble those that characterize diatom assemblages from extreme environments. According to these data, our hypothesis that diatom taxa from different dates representing distinct stages of succession would be found on the blades of E. arborea, together with abundant pioneer taxa or species from advanced stages of colonization, was not supported, although low values of diversity were observed. We conclude that the structure of the diatom assemblage is atypical, similar to those of assemblages from extreme environments, because of the uncommon (very) low diversity values (S, H´, J´). However, structure analysis did not permit the distinction of succession stages in the epiphytic diatom assemblage.

Discussion

Although macroalgae are considered an ideal substratum for many species of diatoms and recent studies in NW Mexico have proven this to be so, in this case the brown alga E. arborea represents a good substratum only for a few diatom taxa, particularly P. inane, but not for most of the recorded taxa in this study. The scarcity of diatom cells and low species richness, along with the pioneer observation showing no diatoms on E. arborea blades (^{Siqueiros Beltrones et al., 2002}) suggest a sui generis substratum whose study requires at least high-frequency observations. The iconographic catalog in this report may prove to be a useful reference when undertaking future studies on the epiphytic diatoms of E. arborea, inasmuch as the frequency of the diatom taxa and their time variation can be monitored in order to infer changes in the epiphytic diatom assemblages.

The observed species richness was approximately half of what has been recorded in similar floristic studies for either a single or for multiple macroalgae species (^{Siqueiros Beltrones, 2002}; ^{Siqueiros Beltrones and Hernández Almeida, 2006}; ^{Argumedo Hernández and Siqueiros Beltrones, 2008}). In contrast, the floristic study mentioned earlier of a single specimen of P. cartilagineum showed it to be heavily epiphytized by many of the 42 taxa recorded, and coupled to a much higher value of diversity (H´= 3.52), a low dominance and a high equitability, indicating that the environment provided by the host is very favorable (^{Siqueiros Beltrones and Argumedo Hernández, 2014a}). Meanwhile, the low values of diversity observed for the diatom assemblages on E. arborea blades resemble those observed in diatom assemblages from extreme environments (^{Siqueiros Beltrones, 2002}). The subtidal habitat is characterized by high energy which causes very dynamic movement of the E. arborea thalli, affecting the settlement of diatoms on the violently moving blades. Colonization is expected to be higher during periods of calm waters.

Because similarity values were high in subsamples from a same sample (^{Siqueiros Beltrones and Argumedo-Hernández, 2014a}), they represent the same structure and ensure that the subsamples did not differ from the overall epiphytic assemblage. This suggests that the examined samples represent but one assemblage that does not show important temporal variations on the basis of species composition and association structure strongly influenced by Pteroncola inane. Accordingly, the abundance of P. inane determines the basic structure of the assemblage.

It is thus strange that, in spite of being so abundant, P. inane set a new record for the region during this investigation (^{Siqueiros Beltrones and Argumedo Hernández, 2014b}). Such monospecific proliferations have been observed on blades of M. pyrifera e.g., Cocconeis costata var. pacifica (Grunow) Grunow (^{Siqueiros Beltrones et al., 2002}) and Rhoicosphenia genuflexa (Kützing) Medlin (^{Argumedo Hernández and Siqueiros Beltrones, 2008}); in these cases, however, the monospecific dominance has not been that extreme.

Based on the above, an expected succession of the taxa in the assemblage could not be observed because, although the abundance of P. inane was highest in the September sample, in the other dates most taxa also remained rare or uncommon. That is, the abundance of cells depended heavily on the proliferation of P. inane. Given the objective of this study, this property (abundance) was not quantified, but it may be inferred from the valve concentration in the examined permanent mounts (Table 2).

As noted above, the initial reported absence of diatoms on E. arborea (^{Siqueiros Beltrones et al., 2002}), and the scarcity of most diatom taxa in the present study merely suggests that distinct stages of succession occur on the E. arborea blades as proposed in our hypothesis, but it precludes its confirmation. It is tempting to suggest that the abundance of Pteroncola inane in the examined samples represents the proliferation of a pioneer species that conditions the blade surface of E. arborea for other taxa, serving as an alternate substrate to E. arborea which could be hostile to the other diatoms. Epiphytism between diatoms is a common phenomenon (^{Round et al., 1990}), and has been documented for this region in epiphytic diatom assemblages of M. pyrifera (^{Siqueiros Beltrones et al., 2002}) and of P. cartilagineum (^{Siqueiros Beltrones and Argumedo Hernández, 2014a}). To contrast this new hypothesis, an ex profeso design is required, including a higher frequency sampling species, i.e., several samplings within a single season to contrast said hypothesis, which may also prove useful in monitoring other diatoms.

From a pragmatic point of view, in spite of the low species diversity of epiphytic diatoms observed on blades of E. arborea, the sole proliferation of Pteroncola inane combined with many rare or uncommon taxa are expected to enrich the kelp´s nutritional value which would favor various species of grazers including those of economic importance like abalones (Haliotis spp.). In fact, much of the interest in the study of M. pyrifera focuses on its role as food for abalone. As with other macroalgae, epiphytic diatoms are considered to be enriching their food value (^{Siqueiros Beltrones and Argumedo Hernández, 2005}; ^{Siqueiros Beltrones and Argumedo Hernández, 2014a}), inasmuch as many epiphytic diatoms have been observed within the gut contents of juveniles and adults of abalone collected in the wild (^{Siqueiros Beltrones, 2002}; ^{Siqueiros Beltrones et al., 2004}; ^{Siqueiros Beltrones et al., 2005}). Thus our intention to determine what diatom taxa would be found on other macroalgae, such as E. arborea, which represents the ecological alternative for M. pyrifera in the region. The assumption that E. arborea blades may be grazed by abalone is supported by recent experiments that show that abalone juveniles from the region successfully feed on E. arborea (^{Mazariegos-Villarreal et al., 2012}).

Knowing the taxonomic identity of the main epiphytic diatoms that are being ingested by Haliotis spp. together with their hosts offers an alternative to a better management of abalone under culture conditions. It should also prompt studies on the nutritious properties of these taxa in order to pinpoint their specific role in the diet of the many grazers that depend on them.