Coral hosts provide more than shelter to boring bivalves

Coral reefs are slow-growing, calcareous, and biodiversity hotspots (see Stella et al., 2011 for review). Bioerosion of living corals may accelerate the loss of coral skeletons and prevent reef accretion (Harriott & Banks, 2002). Despite their impact on coral reef resilience and stability, adequate knowledge of the ecophysiology of bioeroders is lacking (Schönberg et al., 2017). To better understand the nature of the relationship between bioeroding (boring) bivalves and their coral hosts, we conducted field studies over 12 months to compare the physiological parameters of Leiosolenus simplex residing in live coral hosts and in corals stripped of their live tissue (therefore considered "dead"). Surprisingly, 1 year after the experiment, the survival rate of bivalves living within dead coral hosts was similar (<5% difference) to that of the bivalves in live coral hosts, and the diet composition of the bivalves residing in both live and dead coral hosts was quite similar. Nonetheless, several other indicators of physiological condition, such as pumping rate and O2 consumption of bivalves residing in dead coral hosts, were considerably lower (31%) than those residing in live coral hosts. Most bioeroding species are cryptobiotic (Timmers et al., 2021). Large endolithic bivalves constantly keep pace with the outward growth of their coral hosts by sustained outward mechanical and chemical burrowing (Lazar & Loya, 1991; Soliman, 1969), allowing them to maintain contact with the surrounding water. Additionally, both live corals and their endolithic bivalves are suspension feeders, raising the possibility of trophic overlap and competition. Therefore, it may be advantageous for coral-boring bivalves to inhabit dead coral hosts, thereby receiving the benefit of mechanical protection without the energy costs of continuous burrowing or competition resulting from trophic overlap. Indeed, boring into dead substrates is assumed to be the ancestral condition that persists into the present in many coral-boring species (Kleemann, 2008; Morton, 1990). Some species, such as those of the mytilid genus Leiosolenus (formerly of the genus Lithophaga), preferentially or exclusively inhabit living corals (Mokady et al., 1992; Scott, 1988), suggesting that the bivalve may derive essential benefits from this relationship. Such benefits may include active protection provided by the stinging coral cnidocytes (Morton, 1990; Morton & Scott, 1980; T. Amit, personal observations over 10 years of fieldwork) or the consumption of mucus for nutritional needs (Shafir & Loya, 1983). The nature of the relationship of the boring bivalve to its coral host is thus an intriguing question; however, aside from Yahel et al. (2009) and Amit et al. (2023), the extreme inaccessibility of these cryptobionts has seriously hampered research on their ecophysiology. Our field study aimed to elucidate the relationship between the boring bivalve L. simplex (Iredale, 1939) and its coral host species Astreopora myriophthalma (Lamarck, 1816) in the Gulf of Eilat/Aqaba, Red Sea, Israel. We used purpose-built watertight instrumentation and scuba manipulations to compare, in situ, key physiological and trophic status indicators for L. simplex inhabiting 12 similar-sized, haphazardly chosen coral heads sampled in December 2017 (under Israel Nature and Parks Authority permit number 41787). Six of the sampled coral heads were stripped of their live tissue using the PAASCHE airbrush system (Alamaru et al., 2009). The bivalves in the live and stripped coral colonies were allowed to acclimate in the field for 1 month prior to the beginning of the measurements. The oxygen consumption of the bivalves (an indicator of metabolic activity) was quantified in situ, using optodes (FireSting-O2 system, PyroScience, Germany) and an automatic logger (Figure 1a, see Moskovich et al., 2023 for a detailed description of the methodology). The proportion of actively pumping L. simplex was quantified monthly using fluorescein dye released at the inhalant siphon opening of the boreholes (Figure 1b–d, see Video S1 for demonstration). Pumping rates were measured using a modified version of the dye front speed (DFS) method described by Yahel et al. (2005); for a demonstration, see Video S2. To elucidate the putative food sources of the bivalves, we measured the δ13C and δ15N stable isotope ratios of L. simplex tissues and of the particulate organic matter (POM, <100 μm) in the overlying water, using standard methodology (Sharp, 2017). Plastic nets deployed on top of the experimental setup protected the coral colonies from predation. The O2 consumption of L. simplex was markedly reduced in the stripped coral hosts compared with that of L. simplex in the live coral hosts at both 1 and 12 months post-manipulation (Figure 2a). After 12 months, the percentages of actively pumping (i.e., known living) bivalves were relatively high (80%–85%) in both the living and stripped coral hosts. However, the percentage of actively pumping bivalves in the stripped coral hosts remained consistently lower in comparison with that of the live coral hosts throughout the 12-month observation period (Figure 2c). Thus, the reduced percentage of pumping bivalves in the stripped hosts before the 12-month mark did not signify higher mortality, but rather a lower frequency/duration of pumping activity than those in the live hosts. At the 12-month mark, L. simplex residing in live coral hosts pumped at the same rate as those in the dead coral hosts, indicating that the high oxygen removal was not the result of a lower pumping rate but rather of higher oxidation of organic matter indicative of superior feeding (Figure 2a). Given that both the boring bivalve and its coral host are suspension feeders, with the possibility of trophic niche overlap, the obvious next question was "How did removing the live coral tissue affect the diet of L. simplex?" The stable isotope data showed that the diet composition and trophic position of L. simplex were similar in live and dead coral hosts, with nearly identical δ13C values (indicating similar nutritional sources), and an observed δ15N difference ≈1‰, indicating a slight increase in trophic level of diet items in bivalves inhabiting dead corals (a difference of 3‰ signifying a change in trophic level; Adams & Sterner, 2000) (Figure 2b). The slight increase in the trophic level (δ15N) of bivalves inhabiting dead corals supports the possibility that the lack of competition from the coral polyps for zooplankton may be at play (Figure 2b). Although bivalves in both the live and the dead coral hosts showed typical stable isotope fractionation relative to reef POM (Duarte et al., 2018; Hondula & Pace, 2014), which they have been observed to capture (Amit et al., 2023; Yahel et al., 2009), the δ13C signature of the reef POM at the study site was much lighter (−25.1‰) than that of the bivalve tissue (−19.4‰) (Figure 2b). Coral tissue and mucus δ13C values in the Gulf of Eilat/Aqaba range from −10‰ to −19‰ (Alamaru et al., 2009; Muscatine et al., 1989; Naumann, Mayr, et al., 2010), suggesting that some fraction of the bivalve diet was derived from live corals, potentially coral mucus (Marshall, 1968; Naumann, Haas, et al., 2010; Shafir & Loya, 1983). Living within mucus-generating (i.e., live) corals may thus confer a trophic advantage to the boring habit of L. simplex. Coral mucus was likely also indirectly available to the individuals inhabiting the stripped corals in our experiment due to their physical proximity to the live corals (Johannes, 1967), thereby contributing to the similar δ13C values in both groups. Because aerobic respiration is the metabolic pathway for obtaining energy from assimilated food, it appears likely that the difference in O2 consumption signaled an overall reduction in the quantity of food assimilated by the bivalves inhabiting the stripped coral. This interpretation is supported by the pumping rate data, which showed reduced O2 consumption even when the bivalves inhabiting stripped coral hosts had a higher pumping rate than those inhabiting live coral hosts (Figure 2a). Whereas the stable isotope data only inform as to the origin of the food and do not allow a quantitative comparison of the amount of each food type consumed/assimilated, the O2 consumption data argue for a greater amount of food being metabolized by the bivalves in the live coral heads, suggesting superior physiological condition. Taken together, these data show that there were small differences in mortality or diet composition between L. simplex in stripped versus live coral hosts, compared to the larger differences in pumping rate and O2 consumption, during the 1-year experiment. The data indicated that live coral hosts provide a greater food supply, thus providing scope for a greater investment in growth and reproduction (Figure 2a,c). Hence, while it is possible for L. simplex to survive within a dead coral host for an extended period, a live coral host enhances the physiological state of this boring bivalve. It should be noted that over hundreds of survey dives in the natural reefs of the Red Sea, we did not observe any L. simplex living in naturally dead corals (T. Amit, G. Yahel, and Y. Loya, personal observations), underscoring the additional role of live corals in the protection of both the coral host and L. simplex from predation. This study demonstrates the potential for advances in understanding interactions between cryptic reef species by augmenting natural history observations with in situ experimental manipulations using appropriate underwater instrumentation. We plan to use a similar field deployment to examine how live coral hosts influence the reproductive biology and growth of boring bivalves and vice versa. The data presented here suggest that the perks of life within a live coral host go well beyond the provision of shelter. We thank Prof. Maoz Fine for suggesting the field study design, Asaph Rivlin and the entire staff of the Inter-University Institute in Eilat (IUI), the Faculty of Marine Science in Michmoret (Ruppin Academic Center), and the Yahel and Loya lab members for their material and technical support, as well as Rei Diga, Shir Pessach, and Rachel Pardo for help with the lab work and analyses, and Tom Shlesinger and Rei Diga for their underwater photography. Special thanks are due to Prof. John Pastor for his constructive criticism of this manuscript. PGB thanks ISOMer for providing facilities over the course of this study. Funding was provided by the Israel Science Foundation (ISF) grant 640/21 to Yossi Loya and ISF grant 249/21, United States-Israel Binational Science Foundation grants 2012089 and 2017622, and National Science Foundation-IOS grant 1755409 to Gitai Yahel. The authors declare no conflicts of interest. Data (Amit, 2023) are available in Zenodo at https://doi.org/10.5281/zenodo.7965711. Video S1. Video S1 Legend. Video S2. Video S2 Legend. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.