A colony of live massive coral, Porites, was collected from the intertidal zone of Sichon (9°00′16″N, 99°55′20″E). It was then broken into small pieces. There were about 10–15 corallites in each piece (nubbins). A nubbin was attached on top of a plastic rod using a very small amount of ethyl-2-cyanoacrylate glue and then left in the air for several seconds for the hardening of the glue. All coral nubbins were vertically set onto the plastic net (approximate mesh size 5 mm). The plastic net (culture frame) with coral nubbins was hung in the ocean for 2 months. The survival rate was monitored weekly.

Thirty coral nubbins were collected from the culture frame and then were acclimatized in the lab for a week prior to the experiment. There were five treatments of different renieramycin M concentrations: 0.00 ppm (control), 0.02 ppm, 0.20 ppm (natural concentration), 1.00 ppm, and 10.00 ppm. One hundred microliters of dimethyl sulfoxide (DMSO), which was preliminarily tested and showed no effects on Porites, was used as a solvent for each treatment, including the control. The total volume was made to be 50.0 ml by 30 ppt of filtered seawater with natural salinity. For each treatment, 5 coral nubbins were submerged in the solution with air bubbles for 8 h. After that, coral nubbins from each treatment were separately placed into transparent containers, each of them containing 500 ml of seawater. To control the growth of algae during the experiment, every container was left under approximately 40% of natural light. Bleaching and necrosis were observed daily for 2 weeks.

This method was modified from Maida et al. (2006). Several dead coral rocks of Porites sp. with approximately 20 × 30 × 20 cm³ (W × L × H) were collected from an intertidal zone of Sichon (9°00′16″N, 99°55′20″E). They were then cut into approximately 4 × 4 × 1 cm³ (W × L × H) cubes. Each coral cube was desalinated by submerging it in distilled water several times. After that, the coral cubes were dried in an oven at 50°C for 2 days. In all of the cubes, a hole (6.35 mm in diameter) was made at the center by a drilling machine.

Five experiments were prepared with 0.00 ppm (control), 0.05 ppm (natural concentration), 0.10 ppm, 0.20 ppm, and 2.00 ppm renieramycin M in 500 ml of methanol. Three cubes were randomly assigned for each experimental unit. The cubes were submerged into the methanolic solutions for 24 h. Then, they were hung and dried at room temperature (28°C). The experiment contained four experimental sets. Each set consisted of a control unit and four treatment units with different renieramycin M concentrations as described above. There were three replicates of each unit. All experimental sets were then hung in the sea (approximately 1 m below the low tide level) at the same coral reef habitat where the dead coral rocks were collected ( Figure 2 ). Each replicate was approximately 10 cm deep apart from each other. Thus, the depths did not have any effects. The first experimental set was collected from the sea 5 days after the experiment started and the second, third, and fourth experimental sets on days 10, 15, and 20, respectively. Only the upper and lower surfaces of every cube were investigated for fouling organisms under a binocular stereoscope. The number of sessile organisms was counted and grouped into class taxa level. Species diversity index (H’) was used to determine the fouling effect of renieramycin M as well.

Fresh seawater, collected from a coral reef habitat from Sichon (9°00′16″N, 99°55′20″E), was aerated 24 h prior to the experiment. Then, 1.00 ml of this water was pipetted using a micropipette and was poured onto an aerobic count plate (petrifilm sheet) (3M Petrifilm Aerobic Count Plate, 3M) for each treatment. A petrifilm sheet was unmoved for a couple of minutes to let the gel solidify. Paper disks (5.0-mm diameter) (MF-Millipore Membrane Filter, Merck) have 200, 100, 10, or 1 µg of renieramycin M. As a control, paper disks bearing 0.0 µg of renieramycin M were prepared ( Figure 3 ). The petrifilm sheets were placed and incubated at 30°C for 24 h. In one plate, both a control disk and a treatment disk (a single concentration of renieramycin M in each plate) were placed. After 24 h, the diameter of the zone of inhibition in each paper disk was recorded. The number of bacterial colonies growing on the grid areas containing the paper disks was also counted by using an illuminated magnifier and a hand tally counter.

The differences on C, H, and N contents and renieramycin M concentrations with Porites or Palythoa were examined using the nested analysis of variance (ANOVA) to examine both differences between coexisting species and among the areas, as the factor of the areas is nested in the factor of the coral species. In addition, the numbers of bacterial colonies on paper disks (3 disks for control and 3 disks for treatment) of each treatment petrifilm sheet were statistically compared using t-test. A significant level applied in this study was 0.05.

The results showed that C, H, N contents and renieramycin M concentration were not significantly different (P > 0.05) among areas of Xestospongia coexisting with Porites and Palythoa ( Table 1 ). The nutrition values (C/N ratio) at different areas of Xestospongia coexisting with Porites and of those coexisting with Palythoa were also not significant (P > 0.05) ( Table 1 ). However, 8 out of 15 Xestospongia individuals that coexisted with Porites had their highest renieramycin M concentrations in the edge area, whereas 7 out of 15 Xestospongia individuals coexisting with Palythoa had the highest concentrations in the outer area ( Table 2 ).

Bleaching and necrosis of coral nubbins were not observed during 2 weeks in every treatment. Coral nubbins were still brown in color in both control and treatments and sometimes covered by a mucous sheet. During nighttime or low light intensity, corallites of some coral nubbins protruded and waved their trunks and their tentacles into the water column.

Renieramycin M was found to inhibit the settlement of some sessile organisms. However, the duration of the antifouling effect could be found on the first 5 days ( Figure 4 ). There were three groups of sessile organisms found on the control plates, pelecypoda, polychaeta, and barnacle (Semibalanus balanoides), whereas only pelecypoda was found on the treatment plates during the 5-day period. After 10 days, pelecypods were the major group of sessile organisms on control and treatment plates. Polychaeta was found settled on the treatment plates of 0.05–2.0 ppm concentrations ( Figure 4 ). After a 15-day period, S. balanoides was found again on the settlement plates (control, 0.05 ppm, 0.1 ppm, and 2.0 ppm), whereas at the 5-day period, S. balanoides was found only on the control plates ( Figure 4 ). The composition of sessile organisms on every treatment was the same as the composition of sessile organisms on the control after 20 days ( Figure 4 ). The diversity index also supported the effect on antifouling of renieramycin M ( Figure 5 ). After 5 days, the highest diversity index was for the control plates, whereas other treatment plates had a zero diversity index. After longer periods (days 15 and 20), the diversity index values were high in all treatments.

There was no significant difference (P > 0.05) in the number of aerobic bacterial colonies in all contents of renieramycin M ( Figure 6 ). The average number of bacterial colonies on the control paper disks ranged from 30 to 62 colonies per 50 µl, whereas the average number of colonies on treatment paper disks ranged from 45 to 61 colonies per 50 µl ( Figure 6 ). The zone of colonial growth inhibition was not observed on any control or any treatment paper disks. The statistical analysis of antimicrobial effects between different concentrations of renieramycin M compared with control was shown in Table S1 .

In this study, the C, H, and N contents were uniformly allocated in the body of Xestospongia coexisting with Porites, even though there were no significant differences on C, H, or N contents in same areas of Xestospongia when it coexisted with different competitors (Porites or Palythoa) ( Table 1 ). However, the varieties of nutrition, chemical contents, and morphological and physiological changes were commonly found in plants, sponges, and other sessile organisms (Haukioja et al., 1998; Swearingen and Pawlik, 1998; Puglisi et al., 2000; Furrow et al., 2003; Meroz-Fine et al., 2005). For example, for plant species in boreal forests, plants balance their C/N utilizations to survive in the intensely competitive and herbivory pressured habitats (Bryant et al., 1983). In addition, some N-based chemicals, produced by organisms, were varied by N-limited habitats (Bryant et al., 1983).

There were no significant distributions of renieramycin M concentrations along the body parts of Xestospongia ( Table 1 ). On the other hand, other sponge species, Oceanapia spp. and Crambe crambe, have been proven to significantly distribute their toxicity and secondary metabolites within specimens such as at fistule, capitum, basal, or even different cell types (Uriz et al., 1996; Schupp et al., 1999; Salomon et al., 2001). By comparison, renieramycin M concentrations at the edges were higher in Porites-coexisted Xestospongia than Palythoa-coexisted ones. These results might be due to the allocation of chemicals on different spatial competitors of Xestospongia. Although there were a number of studies on the effects of allelochemicals and secondary metabolites of marine sponges on potential spatial competitors, particularly corals, the comparison of chemicals produced by sponges when they coexist with different potential spatial competitors was less available (see examples in Porter and Targett, 1988; Aerts and Van Soest, 1997; Aerts, 1998; Nishiyama and Bakus, 1999; Engel and Pawlik, 2000; Pawlik et al., 2007).

The effects of allelochemicals of sponges on corals and on other sponge species were necrotic effects, bleaching, and the effect on the photosynthetic ability of coral’s symbionts (Porter and Targett, 1988; Engel and Pawlik, 2000; Pawlik et al., 2007). However, renieramycin M exhibited neither bleaching nor necrosis effects on Porites.

During the surveys, the necrotic scars or the dead zones of Porites were observed only on the tissue that contacted Xestospongia (in this study; Darumas et al., 2007). Therefore, the dead zones or necrotic areas might be caused by the synergy of physical damage (by spicules) and other allelochemicals. In addition, sessile organisms were known to compete for space. The strategies to compete with their competitors include overgrowth, physical damage, allelochemicals, and waterborne substances (Porter and Targett, 1988; Paul, 1992; Aerts, 1998; Nishiyama and Bakus, 1999). According to the field surveys of this study, the interactions between Xestospongia and the massive coral, Porites, were of the single category, sponge overgrowing coral. The sponge/coral interactions were classified into four categories: overgrowth, peripheral contact, tissue contact, and non-contact (Aerts and Van Soest, 1997). Their study showed that 46.6% of Xestospongia spp. (X. caminata, X. muta, X. proxima) had tissue contact with corals, and 32.6% had non-contact with corals, while 2.5% of Xestospongia overgrew corals. On the other hand, just 2.2%, 14.3%, and 33.8% of massive coral, Porites astreoides, had overgrown, peripheral contact, and tissue contact by sponges, respectively. In addition, the major proportion of the interaction (49.7%) of this coral was non-contact to sponges (Aerts and Van Soest, 1997). Although, the effect of overgrowth by sponges on corals was not directly documented, one study showed that the overgrowth of sponge on introduced mussels in the Great Lakes (USA) caused mortality or at least loss in the mussels (Ricciardi et al., 1995).Renieramycin M had no antifouling effect on pelecypods but on polychaetes and barnacles (S. balanoides) during the first 5 days of the experiments. The antifouling chemicals produced by pre-settlement organisms were not universally against newcomers (Porter and Targett, 1988; Paul, 1992; Nishiyama and Bakus, 1999). Antifouling was the active competitive defense of sessile organisms. This strategy prevented interspecific larva from settling on surfaces of the previous one (Jackson and Buss, 1975; Nishiyama and Bakus, 1999; Kubanek et al., 2002). The antifouling process operates through mechanical and chemical mechanisms. Some soft corals were swept out and were damaged by sweeper tentacles and nematocysts of hard corals (Paul, 1992). Some sessile organisms used at least two mechanisms to inhibit interspecific larval settlement (Bak and Borsboom, 1984; Coll et al., 1987; Sears et al., 1990; Henrikson and Pawlik, 1995; Hirota et al., 1998; Nishiyama and Bakus, 1999). First was a waterborne substance production, which was released into the water column to affect the competitor’s fitness. Second was the allelochemical that affected the competitors when the tissues were in contact.

Renieramycins were tested to inhibit Staphylococcus aureus, Bacillus subtilis, and Vibrio angularium but not Escherichia coli, Candida albicans, Pseudomonas aeruginosa, and Enterobacter aerogenes (Frincke and Faulkner, 1982; Kelman et al., 2001). In this study, renieramycin M did not inhibit the growth of aerobic bacteria in any treatment. However, antibiotic chemicals produced by sessile organisms or the symbionts were effective against specific bacteria rather than a broad spectrum (Frincke and Faulkner, 1982; Kelman et al., 2001). A number of studies elaborated more about the interaction between bacteria vs. sponge and bacteria vs. macroalgae such as anti-non-symbiotic species and antifouling (Imhoff and Trüper, 1976; Wilkinson, 1978; Lemos et al., 1985; Althoff et al., 1998; Webster and Hill, 2001; Thacker and Starnes, 2003; Armstrong et al., 2006). Sponges of the genus Dysidea were a specific host of cyanobacterium (Thacker and Starnes, 2003). Sometimes, the chemicals produced by microsymbionts were beneficial to the host (Lemos et al., 1985; Webster and Hill, 2001). Epiphytic bacteria attached to seaweed produced antimicrobial chemicals for the host (Lemos et al., 1985). α-Proteobacteria, surrounding the choanocyte chamber of Rhopaloeides sponge, was expected to play the role of nutrient uptake in the host sponge (Webster and Hill, 2001). In some cases, microorganisms may be harmful to sessile organisms. This can be inferred by the varieties of antibiotic substances produced by sessile organisms or their symbionts (Frincke and Faulkner, 1982; Lemos et al., 1985; Kelman et al., 2001; Kubanek et al., 2002; Müller et al., 2004).