Epiphytic diatom assemblages on invasive Caulerpa taxifolia and autochthonous Halimeda tuna and Padina sp. seaweeds in the Adriatic Sea – summer/autumn aspect

The marine green macroalga Caulerpa taxifolia (M.Vahl) C.Agardh has pinnate, fern-like fronds (up to 25 cm long, 2 cm wide) that extend upward from horizontal stolons (Meinesz et al. 1995; Fig. 1). This alga outcompetes native seaweeds and seagrasses in the Mediterranean by forming dense carpets, leading to a loss of biodiversity. It is indigenous to tropical and subtropical seas worldwide, including Australia (Phillips & Price 2002) and has been widely used as a decorative plant in the marine aquarium trade. It was accidentally released from the Monaco Aquarium in 1984 (Meinesz & Hesse 1991), rapidly spread across the western Mediterranean basin (Meinesz et al. 2001) and became one of the most invasive species. In the eastern Adriatic, C. taxifolia was observed for the first time in 1994 in the Bay of Stari Grad (43°10’54”N, 16°35’00”E) on a hard, sandy and muddy substrate without vegetation or within meadows of Posidonia oceanica (L.) Delile (Žuljevic & Antolic 2002).

Figure 1

Caulerpa species are characterized by the presence of secondary metabolites, such as caulerpenyne (CYN), the main function of which is as a chemical defense mechanism against herbivores and epiphytes (Box et al. 2008; Sureda et al. 2009). The maximum concentrations of CYN for the “aquarium-Mediterranean” strain of C. taxifolia were recorded in autumn and the minimum in spring, reaching values that are much higher than those observed in other Caulerpa species (Dumay et al. 2002). Caulerpa taxifolia has a marked seasonal biomass cycle with higher biomass corresponding to higher water temperatures (Meinesz et al. 1995). In summer and autumn, C. taxifolia can grow by nearly 2 cm per day (Meinesz 2002) and during this period fronds of C. taxifolia reach their maximum length (Meinesz & Hesse 1991; Meinesz et al. 1993; 1995). As the growth and toxicity of C. taxifolia vary greatly throughout the year (Amade & Lemée 1998; Thibaut et al. 2004), the chemical defense of this species may affect the settlement and development of different sessile organisms in invaded systems to varying degrees (Prado & Thibaut 2008). As the species turnover is known to be strictly controlled by seasonal variables of the host plant, such as leaf growth (e.g. Wittmann et al. 1981; Prado & Thibaut 2008), it is important to analyze epiphytic diatom assemblages throughout different seasons, which in the case of C. taxifolia means particularly in summer and autumn.

The specific composition of diatom communities of C. taxifolia has not been thoroughly examined, despite the potentially important role of epiphytic diatoms in the functioning of ecosystems influenced by C. taxifolia. The only studies of benthic diatoms from areas of invasive C. taxifolia in the Adriatic were those related to epiphytic diatoms of C. taxifolia and focused on the morphology and description of valve ultrastructure of a new marine diatom, Cocconeis caulerpacola (Car et al. 2012), and the study of epilithic diatom communities from areas of invasive Caulerpa species (Car et al. 2019). The analysis of epiphytic diatoms of another taxa from the genus Caulerpa, Caulerpa racemosa, on the Pacific coast

of Japan was primarily focused on the morphology of taxa from genus Cocconeis (Suzuki et al. 2001). In the Mediterranean Sea, the genus Cocconeis was thoroughly researched when a fine-scale analysis of diatoms associated with Posidonia oceanica was carried out to investigate the composition and diversity of epiphytic communities, with special reference to the most common and abundant genus – Cocconeis (Majewska et al. 2014). The epiphytic diatom communities of the endemic Mediterranean seagrass Posidonia oceanica are among the most frequently analyzed (Mazzella et al. 1994; De Stefano et al. 2000; Majewska et al. 2014). Unlike the Posidonia communities, there are few studies of diatom communities associated with the green macroalga Padina sp. and the brown macroalga Halimeda tuna (J. Ellis & Solander) J. V. Lamouroux in the Mediterranean (e.g. Belegratis & Economou-Amilli 2002).

In general, the majority of previous studies on benthic diatoms in the Adriatic were conducted in the northern Adriatic and estuaries, and seasonal fouling by diatoms was studied on artificial substrates (e.g. Buric et al. 2004; Totti et al. 2007; Caput Mihalić et al. 2008; Levkov et al. 2010; Mejdandžić et al. 2015; Nenadović et al. 2015). Limited information is available on the composition of diatom assemblages growing on either natural or artificial substrates in marine coastal waters of the eastern Middle and South Adriatic, e.g. diatoms from stones in the oligotrophic Bay of Neum in Bosnia and Herzegovina (Hafner et al. 2018a,b) and along the Albanian coastal wetlands (Miho & Witkowski 2005).

The main objective of this study was to describe the dynamics of diatoms on the invasive macroalga Caulerpa taxifolia during a monthly sampling over summer and autumn and to compare the structure of epiphytic diatom assemblages of the investigated invasive macroalgae with the taxonomic composition of epiphytic diatoms on autochthonous brown (Padina sp.) and green (Halimeda tuna) algae in the area inhabited by C. taxifolia in the Adriatic. Due to differing thallus architecture, it was expected that the taxonomic composition of epiphytic diatoms on the investigated macroalgae would vary. In addition, the diversity of dominant diatom genera found in this area is thoroughly described. The results of the analysis of the epiphytic diatom community and its succession on macroalgae on a fine time scale in an area affected by invasive C. taxifolia in the Adriatic Sea are described for the first time.

Altogether, 137 diatom taxa belonging to 42 diatom genera were identified as epiphytes on the upper 2 cm part of C. taxifolia fronds. The genus with the largest number of taxa was Mastogloia (29), followed by Amphora (18), Nitzschia (14), Cocconeis (9) and Navicula (7).

The number of taxa on C. taxifolia fronds increased from June to August, with the smallest number in June (41) and the largest one in August (88). The largest number of taxa on Padina sp. was observed in September (82; Fig. 3a).

Figure 3

The widest (3.11–4.88) range of the Shannon– Wiener Diversity Index was determined for C. taxifolia, with the maximum in August and the minimum in October (Fig. 3b). The reason for the low value of the Shannon–Wiener Diversity Index in June and October 2010 was the high RA (> 40%) of the genus Cocconeis that occurred with a small number of taxa. Cocconeis molesta var. crucifera and Cocconeis scutellum

var. scutellum, which were initially the main fouling components occurring as patches, declined in July to be replaced by Cocconeis caulerpacola, the latter species being dominant in autumn (Fig. 3c). Due to the high RA of C. caulerpacola in September, the maximum abundance of taxa observed on C.taxifolia that month was replaced by the maximum abundance of taxa on Padina sp. The maximum RA of C .caulerpacola of over 40% was observed in October. No correlation between the abundance of C. caulerpacola and sea water temperature was confirmed (Fig. 4).

Figure 4

In June, the taxa richness on Padina sp. and H. tuna was higher than on C. taxifolia and in both cases the value of the Shannon–Wiener Diversity Index was almost the same and amounted to about 4.5. In July, the taxa richness of both autochthonous algae declined (Fig. 3b). While the taxa richness on H. tuna remained unchanged throughout the autumn, the taxa richness on Padina sp. increased in August and the Shannon–Wiener Diversity Index remained at the highest level observed in this study (4.95–4.96) over the next months.

Table 1 lists 22 dominant taxa, which were most abundant in the epiphytic diatom assemblage on C. taxifolia fronds. They dominated in terms of both frequency of occurrence (> 40%) and RA in the total number of samples (> 1%). When combined together in a single sample, these 22 dominant taxa contribute from 70 to 87% of RA. The following taxa, characteristic of summer and autumn on C. taxifolia fronds, occurred in all five samples with an average RA>1%: Cocconeis caulerpacola, Hyalosynedra laevigata, Cocconeis molesta var. crucifera, Cocconeis scutellum var. scutellum, Nitzschia lanceolata var. minima, Nitzschia panduriformis var. panduriformis, Mastogloia crucicula var. crucicula, Mastogloia crucicula var. alternans, Toxarium undulatum, Ardissonea fulgens and Nitzschia angularis.

Colonization of diatoms on C. taxiofolia fronds in summer indicated, however, that taxa of the genus Mastogloia belong to the dominant taxa in the fouling community. In July and August 2010, Mastogloia spp. occurred with high RA (> 20%) and with a large number of taxa, thus contributing to the high value of the Shannon–Wiener Diversity Index. Of the 19 Mastogloia taxa identified in July 2010, the most abundant was Mastogloia crucicula var. crucicula, followed by Mastogloia crucicula var. alternans and Mastogloia cuneata (Table 2, Fig. 5). Although Mastogloia binotata occurred in all five samples of Caulerpa taxifolia fronds, with the highest RA (1%) observed in July, the average abundance was only 0.6% and therefore M. binotata is not included on the list of the dominant taxa of C.taxifolia fronds. In June 2010, only three taxa of Mastogloia, i.e. M. crucicula var. crucicula, M. crucicula var. alternans and M. cuneata, together accounted for 73% of the total abundance of Mastogloia taxa, thus contributing to the low value of the Shannon–Wiener Diversity Index.

Figure 5

The distribution of the relative abundance of the dominant genera over time showed that the majority of these genera reached the high RA in July and August (Figs 6a, 7). The exception was Cocconeis with the lowest RA in July and August and the highest value in October (Fig. 6c). The average abundance of Hyalosynedra declined from 25% in June to 8% in October, whereas Amphora showed an increase in the abundance at the beginning of autumn. No significant difference in RA of Navicula taxa was observed during this study period.

Figure 6

Figure 7

The analysis of the succession of diatoms colonizing C.taxiofolia fronds in summer and autumn by LM and SEM (Fig. 8) showed that colonial tube dwelling forms, e.g. Berkeleya taxa, belonged to the dominant taxa of the fouling community. In the case of the Berkeleya taxa, an interesting switch in the species composition was observed in August and September. The relative abundance of Berkeleya scopulorum, initially the only fouling component representing Berkeleya, decreased in late summer and was replaced by Berkeleya rutilans in early autumn.

Figure 8

The most striking differences in the species composition of the epiphytic diatom community were observed between invasive C. taxifolia on the one hand and the autochthonous green and brown macroalgae on the other (Fig. 9). The cluster analysis performed on the species RA data revealed two groups of significantly different assemblages of epiphytic diatoms inhabiting different hosts (group 1 – Halimeda tuna and Padina sp. and group 2 – Caulerpa taxifolia). The similarity between the two groups, one group consisting of samples of invasive Caulerpa and the other group consisting of autochthonous algae, was less than 40%.

Figure 9

As revealed by the cluster analysis, the type of macroalgae has a greater impact on the diatom community structure than monthly changes.

According to the cluster analysis, the shorter initial period (June–July) was clearly separated from the later sampling (August–September–October) for invasive C. taxifolia and Padina sp. No separation related to the month of sampling has been demonstrated for H. tuna.

In addition to clusters 1 and 2, SIMPROF analysis also distinguished sub-clusters A, B, C, and D. Sub-cluster A contained diatom samples of H. tuna, sub-cluster B contained samples of Padina sp. Cluster 2 contained C. taxifolia samples from June and July (sub-cluster C) and samples from August, September and October (sub-cluster D). The highest similarity was observed between samples of Padina sp. from August, September and October.

The SIMPER analysis identified the group of taxa contributing the most (cumulatively 51%) to the variance between invasive and autochthonous algae: Cocconeis caulerpacola, Navicula ramosissima, Hyalosynedra laevigata, Nitzschia fusiformis, Cocconeis molesta var. crucifera, Cocconeis scutellum var. scutellum, Cocconeis cf. scutellum, Nitzschia angularis, Navicula arenaria var. arenaria, Mastogloia decipiens and Mastogloia pusilla var. pusilla (Table 3).

Altogether, 158 taxa from 46 genera were identified as epiphytes on the investigated brown macroalga Padina sp. between June and October 2010. Genera with the largest number of taxa were Mastogloia (39), Amphora (21), Nitzschia (13), Diploneis (12), Licmophora (7) and Cocconeis (7). The diatom assemblage of the brown alga was characterized by high RA of taxa belonging to Mastogloia (Table 4). Of the 30 dominant taxa, 12 taxa represented Mastogloia. The average abundance of M. binotata, M. crucicula var. alternans, M. cuneata, M. decipiens, M. ignorata, M. pseudolatecostata and M. pusilla var.

pusilla was higher than 2%. The average abundance of M. decipienson Padina sp. was 4.5%, reaching up to 8.4% in August 2010. The diatom assemblage on Padina sp. in June and July 2010 was dominated by Nitzschia fusiformis with RA of 13.48% and 26.25%, respectively. In August, September and October, diatom assemblages were dominated by Navicula ramosissima with RA of 11%, 7.5% and 13% respectively.

In general, epiphytic diatom assemblages of Halimeda tuna are different from those hosted by Padina sp. As revealed by the SIMPER analysis, Cocconeis scutellum var. scutellum, Nitzschia fusiformis, Navicula ramosissima, Navicula sp.1, Nitzschia angularis, Mastogloia decipiens, Amphora acutiuscula, Mastogloia pusilla var. pusilla, Mastogloia cuneata, Mastogloia ignorata, Rhopalodia pacifica, Tabularia ktenoides, Hyalosynedra laevigata and Berkeleya rutilans contributed the most (cumulatively 51%) to the variance between assemblages (Table 5).

Altogether, 115 diatom taxa representing 34 genera were recorded as epiphytes on H. tuna from the Hvar sampling site. H. tuna hosted diatom assemblages dominated by taxa belonging to Mastogloia (20), (C. taxifolia, Padina sp., Halimeda tuna). The highest Global R (0.85) value was obtained in the analysis of similarity between diatom communities associated with the type of different macroalgal species (invasive/autochthonous; Table 7). The same test performed on the species RA data indicated that the difference between diatom communities over different sampling seasons was not significant (p >0.05). In terms of growth forms, no significant differences were found between the diatom communities on different macroalgal host species (Table 7).

Canonical Analysis of Principal Coordinates CAP (Fig. 10) showed that the samples from sub-cluster A (H. tuna) correlated with Navicula ramosissima and Nitzschia angularis, samples from sub-cluster B (Padina sp.) with taxa of the genus Mastogloia (M. decipiens and M. pusilla var. pusilla) together with Nitzschia fusiformis, while samples from sub-cluster D (C. taxifolia; August, September and October) correlated with the abundance of Cocconeis caulerpacola and Cocconeis cf. scutellum.

Figure 10

Amphora (19), Nitzschia (13), Cocconeis (11) and Navicula (8). Of the 30 dominant taxa, six taxa represented Amphora (Table 6), of which Amphora acutiuscula showed the maximum RA of 6% in July. The diatom assemblage on H. tuna in July 2010 was also dominated by N. ramosissima and C. scutellum var. scutellum with RA of 10% and 23%, respectively. In September, diatom assemblages were dominated by the genus Navicula.

ANOSIM indicated that the diatom communities differed significantly depending on the substrate

Due to the scarcity of information on benthic diatoms in the Adriatic, particularly from the area affected by the invasive macroalga Caulerpa taxifolia, the results of this study were compared with the literature data from different seas worldwide and the comparison was not limited to the investigated host (macroalga).

The data in this study confirmed the expected differences in the taxonomic composition and abundance of epiphytic diatoms among the investigated macroalgae. This may be explained, at least partly, by different physical, chemical and biological requirements of hosts and their ecological conditions.

In this study, Cocconeis taxa were the dominant epiphytic diatoms on C. taxifolia fronds. This is consistent with the findings of Majewska et al. (2014) for Posidonia oceanica leaves in the Mediterranean. In the case of the latter host, C. scutellum was the dominant taxon. Tanaka (1986) found that diatom taxa having higher adhesive strength (Cocconeis, Achnanthes) may be abundant on various macroalgae and tolerate adverse hydrodynamic conditions such as wave action.

In addition, Al-Yamani & Saburova (2011) showed that the epiphytic diatom assemblage on intertidal macroalgae, including Padina sp. from Kuwait, was largely dominated by Cocconeis taxa and constituted a continuous, almost monospecific layer. In our case, despite the high taxa richness on Padina sp., the epiphytic diatom C. caulerpacola was absent on specimens of this taxon, while it was frequently found on fronds of C. taxifolia.

Epiphytic diatom assemblages on Padina sp. were different from those hosted by Halimeda tuna. Although this study showed the presence of a smaller number of diatom taxa growing on H. tuna compared to Padina sp., the number of the analyzed samples (three on H. tunavs and five on Padina) could not be ignored. In the case of H. tuna, the Amphora taxa could be considered as frequent. The diversity of Mastogloia (>39 taxa) on Padina was much higher than on the two other macroalgae studied, and with the high frequency of occurrence the taxon significantly contributed to the high value of the Shannon–Wiener Diversity Index.

In this study, the dominant taxa on C. taxifolia fronds, in addition to Mastogloia taxa, were those characterized by the high motility (e.g. Navicula and Nitzschia), capable of moving on the substrate to find optimal conditions. Totti et al. (2009) concluded that adnate diatoms, including mostly Cocconeis taxa, appear to be more affected by the structure of the host surface and their abundance increases on thalli that offer a more complex microarchitecture for colonization, while they were absent in soft thalli with smooth surface. Monoraphid Cocconeis tightly adhere along their raphe, but as organisms capable of slow motility they are also able to migrate to areas with the greatest availability of light and nutrients or the least chance of desiccation due to tidal fluctuations (Mitbavkar & Anil 2004), which are generally the areas that provide the most suitable and productive environmental conditions (Edgar & Pickett-Heaps 1984). The chemical defense of C. taxifolia may reduce the settlement rate and the development of sessile organisms in invaded systems (Prado & Thibaut 2008). Preliminary studies of secondary metabolites from macroalgae showed biological activity against epiphytes (Phillips & Towers 1982; de Nys et al. 1991; 1995). Organisms can produce different types and rates of metabolites such as wastes, nutrients and toxins, depending on the season, grazing pressure, developmental stage and biological cycles of individuals. Production and exudation may also vary among different organs of the same individual. Successful colonizers must have either a large tolerance range or settle during phases when – or on organs where – the composition and quantities of exudates are not harmful (Wahl 1989). For the Mediterranean strain of C. taxifolia, the maximum concentrations of caulerpenyne were recorded in autumn and the minimum in spring (Dumay et al. 2002). In our case, high summer and autumn values of toxic substances could be the cause of reduced diversity of diatoms and contribution of C. caulerpacola. In fact, this can be explained by the targeted chemical defense of macroalgae against particular taxa.

Defense and epibiosis may vary among different macroalgae parts. Jennings and Steinberg (1997) reported the greatest epiphyte abundance on the oldest tissue and the lowest on the youngest one. They also found that the epiphyte abundance was strongly correlated with the age of different parts of the thallus of the sublittoral kelp Ecklonia radiata (C.Agardh) J.Agardh. In the case of C. taxifolia, despite the fact that it grows nearly 2 cm per day throughout the summer and autumn (Meinesz 2002), it seems that C. taxifolia grows simultaneously with C. caulerpacola.

The research of Nenadovic et al. (2015) on 11 different artificial substrates exposed to the marine environment in the eastern Middle Adriatic in spring and on shoots of P. oceanica showed differences in quantitative and qualitative composition of diatoms among newly colonized surfaces and implied the preference of diatoms for specific substrates. Nitzschia, Cocconeis and Navicula taxa, which can be considered pioneer epiphytic diatoms regardless of the colonized substrate (Railkin 2004; Nenadovic et al. 2015), were frequently found on all substrates alongside the dominant diatom taxa, Cylindrotheca closterium and Amphora. Our results are consistent with those of Nenadovic et al. (2015) and Navarro et al. (1989), who also found abundant taxa of the genera Mastogloia, Navicula and Nitzschia on diverse marine substrates (organic or inorganic) in the tropical region of Puerto Rico. In addition, research on P.oceanica leaves showed that the diatom community structure is also affected by the seasonal cycle of the plant, the depth of the meadow, the age of leaves, and the grazing pressure exerted by herbivores (Mazzella 1983; Mazzella & Spinoccia 1992; Mazzella et al. 1994; De Stefano et al. 2000).

The results of the presented study clearly showed differences in the diatom species distribution pattern between the selected autochthonous macroalgae and invasive C. taxifolia. This may be explained, at least partly, by varying thallus architecture of the studied species, given that Padina sp. and H. tuna have calcified thalli. In addition, multiple environmental factors reflecting the seasonal and site-specific conditions could also be manifested in the structure and species composition of algal communities. For this reason, the results of the present study imply only that an epiphytic diatom community is affected by macroalgae, especially those containing toxins. The lack of samples of younger parts cut from indigenous algae does not allow us to draw definite conclusions about succession and seasonal specificity or distribution of epiphytic diatoms on macroalgal hosts. It can be assumed that differences between invasive and autochthonous algae would also be noticeable at an early stage of life and exposure of macroalgae to benthic diatoms. Obviously, in order to increase the accuracy of predictions, further long-term studies of dynamics of epiphytic diatoms on different macroalgae and other substrates, as well as effects of various environmental factors on the growth of epiphytic diatoms are necessary.

In summary, this study presents the epiphytic diatom species composition on the invasive macroalga Caulerpa taxifolia distributed along the coast of Croatia and the effects of lifespan and seasons on diatom assemblages. The diatom community on C. taxifolia fronds is dominated by Cocconeis taxa. The succession of diatom assemblages showed remarkable variations over the months of summer and autumn. We confirmed the hypothesis of host specificity of Cocconeis caulerpacola in relation to Caulerpa by showing the total absence of C. caulerpacola on Halimeda tuna and Padina sp. collected at the same time and at the same sites as in the case of the invasive Caulerpa. Knowledge of the diatom community structure in these areas is important for studies of toxic effects of the host and enables a better understanding of the functioning of ecosystems dominated by invasive C. taxifolia.