(a)

**Table 2.** Results of 2-way ANOVA analyzing the effect of season (summer and winter) and rock type (limestone, schist, basalt, and granite) on the mean percent survivorship of zygotes after treatment by a wave. Survivorship is calculated as a percentage based on the mean zygote density of control plates in each rock and surface type. Data were arcsine-transformed before analysis. There were 5 replicates of each rock and surface type combination.


#### **4. Discussion**

Zygote attachment and survivorship following a wave were greatest on limestone, least on granite, and intermediate on schist and basalt. A wave dislodged some initial settlers, but it did not change the general patterns of initial attachment we recorded. While we did not statistically compare natural and smooth surfaces (experimental design prohibits this), initial attachment patterns were the same between the two surfaces, and patterns of survival were very similar. These results suggest that minerology, fine-scale characteristics of the rock types, or some other characteristic of rock type, *and not contour* (natural vs. smooth), accounted for the differences we observed.

We examined the initial attachment phase, which may not indicate the effect of bedrock type on zygotes attached for longer periods [16]. The exact timing of the transition from initial mucilaginous adhesion to permanent rhizoidal adhesion is not clear. Initial attachment may occur about one hour after fertilization [43], while another study [54] reported that initial attachment occurs from 3 to 6 h after fertilization and refers to the time after fertilization as the 'pre-adhesive' stage. The zygote undergoes an adhesive maturation phase from 7 to 9 h after fertilization [54], which involves the hardening or gelling of the polysaccharide [3,44]. The chemical composition of the adhesives involved in the attachment of *Fucus* zygotes is largely unknown [54,60]. There is evidence, however, that polysaccharide– protein complexes are involved [44], and that two polysaccharides, an alginic acid complex and sulfated fucan, and polyphenols make up the adhesive secretion [54,61]. In our study, we had an initial attachment of up to 120,000 per m<sup>2</sup> , and approximately 25% were able to maintain adherence after wave exposure. This indicates that even if 3–6 h may be termed a 'pre-adhesion' stage, the shorter time used in our experiment is certainly enough for good settlement and persistence following a wave.

#### *4.1. Surface Roughness*

The initial attachment of the nonmotile zygote is, in part, dependent on the physical characteristics of the substrate. The most implicated substrate characteristic in earlier studies of settlement and attachment of marine organisms is surface roughness, or contour [34–37,62]. Surface roughness has been quantified with the development of the engineering roughness index, a dimensionless parameter that relates the proportion of the surface that is recessed and the amount of freedom a spore has to move based on surface topography [42,63]. Zygotes are likely to find refuge from turbulent flow in rock substrates

that are rough enough to offer cracks and crevices that protect germlings [35,44]. Such crevices may be present within a rock type, or between juxtaposed rock types. Most studies on the effects of surface roughness on settling behavior have used artificial substrates (but see [42,64]). Typically, roughness is examined by cutting or drilling grooves or pits of varying size in otherwise homogenous substrates [16,34,62,64,65], by attaching silica grains of different sizes to a homogenous surface [31,49], or by producing a range of rough surfaces using varying grades of abrasive materials on otherwise smooth materials [66]. These results are helpful in indicating what is possible and, therefore, what can and cannot be expected of larvae (as in these studies), or algal propagules [67]. Although Caffey [68] argues that artificial substrates provide little insight as to the factors influencing attachment in the natural environment, they can, if consistent with field observations, provide valuable corroborative data on control mechanisms.

We did not statistically compare natural and smooth surfaces for the initial attachment or survivorship, because plates of different surface types were held in separate trays (i.e., there was a lack of interspersion of treatments). Settling trays were identical and settlement occurred simultaneously under identical conditions, so it is unreasonable to expect a significant tray effect. Furthermore, analyzing survivorship instead of settler density removes much of the possible variability in the concentrations of the zygote solutions added to each tray. Nevertheless, comparisons between surface types are qualitative.

There were few differences in our study in patterns of initial attachment or survivorship following a wave between plates with natural and smooth surfaces (Figures 1 and 2). With both surface types, limestone, which has the finest grain size and least inherent surface roughness, proved to be most suitable, and granite, which has the largest grain size of the rock types tested, the least suitable for zygote attachment. As this pattern persisted for both natural and smooth surfaces, a factor other than the contour of the surface must explain the variation among rock types.

Although topography on the scale of centimeters to tens of meters is important in assessing intertidal environments and community structure, the effect of substratum microheterogeneity, or texture must also be addressed to investigate settlement processes. For a range of taxa, a microtopography smaller than the length of an organism's attachment point generally reduces settlement [42]. Algal spores and zygotes are typically about 5 µm in size, so the most appropriate scale to measure the surface texture of substrates in algal settling experiments may be the grain size of minerals [64]. Mineral grains are the units that are cleaved from rocks as a result of physical weathering. This suggests that grain size can dictate the surface texture in natural environments because the weathering of rocks with large grains will expose larger mineral surfaces than those with small grain size, although it may be difficult to be so conclusive about this parameter for rock types with heterogeneous mineralogic composition (e.g., two-mica granite). Surface heterogeneity on this scale has been shown to affect community structure in many species [3,35,43,44]. Although on a large scale, the natural surface of the rocks used in this study can be classified in order of increasing roughness (limestone-basalt-schist-granite), this may not be the appropriate scale for our study. Surface roughness as it applies to zygotes has been defined as the number of surface planes of the substrate encountered by the zygote [44,62]. Zygotes may experience more surface planes on a rock with small grain size, and fewer on a rock with large grain size. Thus, the scale of roughness as it applies to zygotes may be the converse from that determined on a large scale. Future studies should investigate differences in roughness among natural substrata on scales relevant to zygotes.

The limestone used in this study was very fine grained [53], with grains smaller than the typical size of the zygotes used. Such substrata would present a greater number of surface planes for mucilaginous adhesion than would a substrate with large grain size (i.e., granite). This classification would also be relevant for surfaces cut with a diamond blade. The large mineral grains of granite (due to slow cooling of parent magma) and schist (due to recrystallization of marine sediments) were large enough so that individual mineral grains were actually cut, leaving a smooth surface, while the small grains of

basalt and limestone remained intact. Thus, any advantage in attachment strength due to grain size would persist even if the natural rock surface were cut, which is what we observed in documenting a few differences in the pattern of survival between natural and smooth surfaces (Figure 2). This hypothesis is consistent with the results of one study [64] in which the settlement of barnacle cyprids was enhanced on fine-grained natural rock plates and inhibited on course plates, even though all settlement plates were machine-cut and polished. Our result, however, does not agree with more recent work that suggests that features the same size or smaller than the settling organisms inhibit settlement [69]. Scardino et al. [69], however, found that for very small motile propagules on the order of 7 µm, the effect of attachment points was weak, so the relationship between surface features and sizes of settling propagules may not apply to *Fucus* spores.

It is important to note that while we found no differences between smooth and rough plates in the pattern of initial attachment of zygotes, our results may not be easily extrapolated to the field. In the field, roughness exists on many scales that we did not test in our study. These roughness differences are due in part to rock minerology, wave energy, crystal size, geological processes juxtaposing rock types, and erosional history, and can well influence very local flow patterns and potential attachment angles—and thus, settlement/recruitment. This should be considered in field experiments that may follow from our study.

#### *4.2. Chemical Interactions between Adhesives and Substrata*

As discussed above, there is evidence that a polysaccharide chain of alginic acid comprises one of the initial attachment adhesives in the Phaeophyta. Alginic acid is a linear 1,4-linked block copolymer comprised of beta-D-mannuronic acid (M) and alpha-Lguluronic acid (G) residues [3]. The strength of polysaccharide gels such as alginate are increased by the binding of Ca2+ [3]. Due to differences in structure, polyguluronic acid has a greater affinity for Ca2+ than for polymannuronic acid. The structure of polyguluronic acid better accommodates insertion of the calcium ion, resulting in cross-linking of the polymer chain and a stiffer gel. Higher concentrations of Ca2+ ions at the surface interface of specific bedrock types may result in more cross-linking of the alginic acid polymer chain, thereby producing a more rigid gel and a stickier adhesive. It is probable that such a difference in Ca2+ concentration exists between limestone and granite because the limestone consists of at least 90% CaCO3, which readily dissociates. The availability of free calcium ions may partially explain a higher initial attachment and survivorship of zygotes on limestone compared to some other rock types.

#### *4.3. Free Energy of Rock Surfaces*

Applied research into fouling by marine organisms has focused on the alteration of potential settling surfaces to prevent adhesion. The most implicated physical–chemical characteristic of substrates is the surface free energy, also referred to as wettability [44]. This is defined as an unsatisfied bonding potential at the surface of a material that results in the greater propensity of that surface to bond to dissimilar particles in the surrounding water [45]. Increased surface free energies have been shown to increase the adhesive strength of algal spores [41,44], change the shape of rhizoids that are produced [43], and enhance the attachment of barnacle and bryozoan larvae [70]. The early settlement of meiospores is best described by water contact angle [41], which is related to surface free energy [41]. As this research is generally applied to the development of nontoxic coatings for use in marine industries, artificial substrates such as synthetic polymers, glass, ceramic tiles, and Teflon have been studied in adhesive comparisons [41,43,44]. The homogenous composition of these surfaces and their predictable matrices make it possible to measure the free energy of these surfaces and then to compare these measurements with strength of attachment.

The modeling of natural substrates such as rock, however, is much more complex than for artificial surfaces due to the heterogenous nature of natural surfaces. The heterogenous mineralogy of natural rock makes surface free energy difficult to measure (Berry, pers. com.). Even the lattice of a simple salt such as NaCl is heterogenous (Na is not the same as Cl), so in introducing the highly variable chemical composition of rocks, it is probable that the surface energy of the substrate will vary over small scales within one sample of any given rock type [71]. Furthermore, there will likely be differences in the free energy of surfaces based on the cleavage of the crystals. Adamson [72] indicated that a clean cleavage of a crystal (i.e., through natural weathering) will have a different and probably lower surface energy than would a ground or abraded surface of the same material.

Differences in surface free energy among substrates are also expected to decrease with prolonged immersion in sea water [70,73]. The adsorption of organic and possibly inorganic molecules can occur on clean surfaces, creating a 'conditioning' film within minutes of immersion in sea water [45,74]. This film can alter physio-chemical properties such as the surface energy of the original surface and the effect of the settlement of algae (see [75] for review). This effect is important to note when comparing short-term laboratory settling experiments using fresh substrata to possible long-term effects of substrates on community composition in the field. Holm et al. [73] concluded that while surface energy may be important in determining initial settlement patterns in some fouling communities, it is probably not a major influence on long-term community development. Conversely, Callow and Fletcher [45] reported that, although surfaces with different original surface energies acquired similar films, differences in attachment persisted after immersion.

No direct measurements of surface free energy were available for comparison in our study. Nonetheless, it is possible that differences in surface charges contributed to the differences in survivorship among our substrates, and this possibility should be explored in future experiments. The extreme contrast between the chemical composition of limestone and granite, for example, make it probable that differences in surface energies of these substrates exist, and that limestone is the more highly charged of the two (Berry, pers. com.). The surface free energy of limestone aggregate is greater than that for granite aggregate [76], which is consistent with this speculation. While the chemical characteristics of aggregates and natural rocks in the field are not expected to be the same, the relative differences might be. The higher settlement on limestone compared to granite we measured would agree with Callow and Fletcher [45] who found that adhesion for a wide range of organisms is higher on surfaces with higher surface energies.

#### *4.4. Other Factors*

Other factors that have been suggested to influence algal and larval attachment are salinity gradients [77], color differences [7], pH gradients [3], substrate hardness [43], and the presence of microbial films [78,79]. The effects of color on attachment have been attributed to differences in thermal properties of dark vs. light substrates [7]. As our experiments were run indoors under diffuse light, color is unlikely to be important (but may certainly be an important contrast between our experiments and settlement under field conditions). A pilot study indicated that there was no detectable difference (to +/− 0.01) in pH at the substrate–water interface after 1.5 h among the four rock types. Such an effect cannot be completely discounted, however, because zygotes may be able to detect much smaller differences in pH than the instrument used could measure [80]. Substrate hardness has been implicated in affecting the adult communities found on rocks because softer rocks (such as limestone) will erode more rapidly than granite or basalt. Hardy and Moss [43] concluded that ephemeral species tend to grow on soft substrates, while perennials occur on hard substrates. Variations in microbial films affect the surface tension of substrates and, hence, attachment [70,81,82]. Microbial films were not likely a factor in our experiments, because settling plates were rigorously cleaned and sterilized before testing, experimental trials were short, and the artificial seawater used contained far fewer micro-organisms than did natural seawater. Finally, Amsler et al. [80] reported that surfaces concentrate several nutrients, which stimulate kelp spore chemotaxis or settlement. It is plausible that a

similar effect may occur in *Fucus*, but the measurement of this effect was beyond the scope of our study.

#### **5. Conclusions**

Our results indicate that the adhesion of *F. spiralis* zygotes is influenced by characteristics of natural rocks other than surface contour and that the general pattern of initial attachment to different rock types we observed persists after the zygotes are exposed to a wave. The microheterogeneity (texture) of substrates based on grain size, the effect of variations in surface charge of natural substrates, and the chemical interaction of the initial adhesives at the substrate interface are suggested as possible factors affecting initial attachment success, consistent with our results of contrasting attachment and survivorship of the zygote on different rock types. The difficulty in assessing characteristics such as free energy in natural substrates and the lack of data on the precise chemical composition of the adhesives produced by settling propagules leave some of these questions open for further study. It is clear, however, that natural substrates present a higher degree of variation in factors affecting settlement than simply different degrees of roughness and that these factors operate within the first few hours of zygote attachment. A further study of this interaction and patterns of settlement among naturally occurring rock types may contribute to a better understanding of attachment of algal spores and to the distribution and dynamics of natural communities on rocky shores. Finally, 26 years have elapsed since our study and, while we do not opine that zygote affinity for different rock surfaces changed as a result of a circa 0.5–1 ◦C increase in water temperature [83], many factors that affect settlement could have changed over this period, warranting repeating these experiments.

**Supplementary Materials:** The following are available online at https://www.mdpi.com/article/10 .3390/jmse9090927/s1, Figure S1: Images of the bedrock types used in the settlement experiments.

**Author Contributions:** All authors contributed to the research idea and experimental design. The experiments were performed by D.C.A. in the laboratory of and with assistance of R.L.V. The first draft and initial data analysis were completed by D.C.A. Subsequent drafts and data analysis were performed by W.G.A.J. and P.E.R. Funding for the research was acquired by R.L.V. R.L.V. read and agreed to publish an earlier version of this manuscript, but poor health prevented him from reading the current version. All authors have read and agreed to the published version of the manuscript.

**Funding:** The research was funded by the NOAA Sea Grant program of the University of Maine (#NA89AADSG020 and #NA16RGO157) and the USDA Hatch funds to Maine Agricultural and Forestry Experiment Station.

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

**Data Availability Statement:** Data are available upon request from the first author.

**Acknowledgments:** We appreciate the use of the Bates College Geology Department rock saw, the use of the Fisheries and Aquaculture Research Group (FARG) facilities, and the assistance of FARG personnel with the seawater system. Comments by an anonymous reviewed significantly improved the manuscript.

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

#### **References**

