Across all sampled localities along the coast of Curaçao, a total of > 500 colonies of Scleractinia and > 70 colonies of Alcyonacea were searched for snails. Three species of Coralliophila were found associated with either Scleractinia or Alcyonacea. Besides C. galea and C. caribaea, a new species, C. curacaoensis Potkamp and Hoeksema sp. nov. was found (described in Appendix).
A total of 690 specimens of C. galea, 139 specimens of C. caribaea and 10 specimens of C. curacaoensis sp. nov. were found on a total of 157 scleractinian colonies and 22 gorgonians (Table 3; voucher numbers RMNH.5004260-5004370). Coralliophila caribaea occurred on both host species groups, whereas the other two Coralliophila species were only found on scleractinians. In 10 out of 157 associations, C. galea co-existed on the same host colony with either C. caribaea or C. curacaoensis sp. nov. (on two colonies). Coralliophila snails were found in association with 20 host species in total, belonging to ten scleractinian genera (Table 3): C. galea occurred on 16 hosts species (Fig. 3), C. caribaea on 12 hosts species (Fig. 4), and C. curacaoensis sp. nov. on a single host species (Fig. 5). In addition, C. caribaea occurred on five alcyonacean taxa (Fig. 4). Coralliophila galea and C. caribaea shared seven host species, whereas C. curacaoensis sp. nov. shared its single host species, Madracis auretenra Locke, Weil and Coates, 2007, with C. galea (Table 3).
Fig. 5. In-situ photos of Coralliophila curacaoensis sp. nov. in association with Madracis auretenra. One snail has its proboscis extended into the mouth of a polyp (c).
Fig. 6. Removed coral tissue of various host coral species underneath Coralliophila individuals: Colpophyllia natans (a); Agaricia agaricites (b-c); Porites furcata (d); Madracis auretenra (e); Pseudoplexaura sp. (f); Eunicea sp. (g).
In case of scleractinian hosts, both C. galea and C. caribaea were usually found at the edge of living coral tissue. On corals with a massive or plate-like growth form, snails were usually found at or underneath the edge of their host (Figs. 3d-e; 3h-i; 3k; 4a; 4d-e). Some snails occurred inside crevices and in between ridges of the host colony, in case of for example C. galea on Pseudodiploria strigosa (Dana, 1846) (Fig. 3l) and C. caribaea on Siderastrea siderea (Ellis and Solander, 1768) (Fig. 4g). On large coral colonies of Orbicella annularis (Ellis and Solander, 1786), C. galea commonly clustered together in groups in crevices between the columns that make up the host colonies, again at the edge of living coral tissue (Fig. 3f). Snails on branching corals were found on the branches themselves, such as C. curacaoensis sp. nov. on M. auretenra (Figs. 3g; 3j; 4b-c; 5).
Snails on alcyonacean hosts were usually found at the base of the colonies (Fig. 4h) but also on the branches, usually on a dead patch (Fig. 4f). Some snails situated underneath the edge of the coral base were almost endofaunal (Figs. 4i-k). Snails sometimes occurred in small aggregations, tightly stacked on top of each other (Figs. 4i-j). Predation on host colonies was visible as damage to the soft surface tissue or its complete removal, leaving bare skeleton behind (Fig. 6). Damage on coral tissue seemed minimal for most host species, despite the occurrence of snail aggregations, except for a few host colonies showing more severe damage (e.g., Fig. 3a). In absence of long-term monitoring of coral colonies, it was unclear whether such damage is indeed the result of predation by Coralliophila spp. or if it was already present when the snails arrived.
Shell dimensionsnext section
Shell length of C. galea and C. caribaea ranged from 3.4 to 38.9 mm and 4.3 to 25.8 mm respectively. In both species, a clear host-associated size structuring of shell length existed (F = 13.6; p < 0.0001 and F = 5.8; p < 0.0001 for C. galea and C. caribaea, respectively (Fig. 7). The mean length of C. galea, snails differed among host species (29 combinations of host pairs, post-hoc Tukey HSD test), compared to a total of for C. caribaea (Tables S2-S3 in Online Supplementary Material 4). C. caribaea (nine pairs of hosts) associated with alcyonacean host species that were larger (based on shell length) than snails associated with scleractinian hosts (including all specimens; F = 23.8; p < 0.0001; Fig. 7).
C. curacaoensis sp. nov. was on average smaller (shell length 2.5-8.9 mm) than C. galea and C. caribaea. However, C. curacaoensis sp. nov. was only found on Madracis auretenra, which was also a host of C. galea. Coralliophila galea individuals associated with M. auretenra (shell length 5.4-9.7 mm) were not different in size compared to specimens of C. curacaoensis sp. nov. (F = 4.07; p = 0.061).
Landmarks could be recorded from a total of 504 out of 631 photographed specimens (60.1% of all collected specimens): 442 shells of C. galea, 55 of C. caribaea and seven of C. curacaoensis sp. nov. (Table 3). Principal component analysis on landmark data of all three species revealed six axes, explaining 71.0% of all observed variance, with an ICC > 0.80 that could therefore be considered repeatable. While overlap in shell shape between species existed, all three species were separated on the first and third PC axis, which explained 31.2% and 10.4% of all variances in shell shape (Fig. 8). On the first PC axis, Coralliophila galea shells separated from both C. caribaea (p < 0.0001) and C. curacaoensis sp. nov. (p < 0.0001). On the third PC axis, shells from C. curacaoensis sp. nov. were also separated from C. caribaea shells (p = 0.016). In total, species identity accounted for 19.5% of all observed variance in shell shape.
Intraspecific variation in shell shape of both C. galea and C. caribaea was high (Figs. 9-10). For C. galea, principal component analysis again revealed five repeatable axes (ICC > 0.80, explaining 62.3% of variance in shell shape). For C. caribaea, four repeatable axes were found (ICC > 0.80, explaining 72.2% of variation) by the principal component analysis. Compared to C. galea, the first two PC axes of C. caribaea explained more of the intraspecific variance in shell shape. Most of the intraspecific variation on the repeatable PC axes of both species was related the shape and relative size of the shell spire. Despite high intraspecific variation, no distinct ecotypes could be distinguished in either species, as all specimens clustered together into one cloud without gaps.
For both C. galea and C. caribaea, all factors and interactions had a significant influence on shell shape (Procrustes ANOVA model; Table 4). Though all tested factors contributed to shell shape, the explained variance in shell shape by any factor was low (R2 < 0.20 for all factors and interactions) and residual variance was high (R2 = 0.777 for C. galea and R2 = 0.429 for C. caribaea).
Host-associated differences in shell shape accounted for some of the intraspecific variation in shell shape of Coralliophila spp. In C. galea, differences in host species explained 6.9% of variance in shell shape (Table 4). Among snails from different hosts species, eleven pairwise differences in shell shape were found (Table S4 in Online Supplementary Material 4), that, even though statistically significant, were subtle, and strong overlap in shell shape existed among snails from different host species (Figs. 9-10).
Shell shape of C. caribaea also differed among host species. Firstly, snails originating from hosts of the order Alcyonacea and snails from Scleractinia differed in shell shape, which accounted for 9.3% of the observed variance in shell shape (Table 4). At the host genus level, two pairwise differences were significant (Table S5).
Depth had a small, though significant, effect on shell shape in both C. galea and C. caribaea (Table 4). On top of an overall effect of depth, and a small host-specific effect of depth was observed in both C. galea and C. caribaea.
Since C. curacaoensis sp. nov. only occurred on M. auretenra, host-related differences could not be assessed. Shell length did not contribute to intraspecific variations in shell shape within this species (F = 2.1; p = 0.070).
Table 4. Factors used in the models of shell shape of both Coralliophila galea and C. caribaea. Shell length was transformed with the natural logarithm; p-values are based on 1,000 permutations.
Allometric patterns were important in determining the shell shape in both C. galea and C. caribaea. In both species, shell length explained minor variation in shell shape (R2 = 0.072; p = 0.001 and R2 = 0.041; p = 0.049 for C. galea and C. caribaea, respectively), and these relationships depended on the host on which snails were found (shell length and host species interaction, R2 = 0.046; p = 0.001 and R2 = 0.133; p = 0.004 for C. galea and C. caribaea, respectively). The presence of such host-associated differences was also implied by the regressions between the CAC and shell length, where the slope of these linear regressions varied among snails associated with different host species (regressions were only done for hosts with ≥ 5 specimens; Fig. 11). Post-hoc tests further confirmed the presence of host-associated differences in allometric patterns in both C. galea and C. caribaea. Predicting the hypothetical shell shape of shells associated with a certain host species for a specific shell length, based on linear regressions of PC scores against shell length, clearly showed these differences (Fig. 12). Allometric patterns were subtler in C. caribaea compared to C. galea. In case of C. galea, both the distance (i.e., amount of change per unit of growth) and correlation (i.e., the direction of change) among vectors of shell length were, after Bonferroni corrections (n = 45), different for six pairs of host species (Fig. 12a; Table S6 in Online Supplementary Material 4). In C. caribaea, the correlation of vectors differed for a single pair of host species, while no differences were observed in the distances among vectors (Fig. 12b; Table S7 in Online Supplementary Material 4).
Fig. 11. Allometric patterns in Coralliophila galea (a) and C. caribaea (b). Common allometric component (CAC) is plotted against shell length (a). Separate regressions for host species with five or more specimens are plotted, symbols on the lines represent the R2-value (first symbol) and p-value (second symbol) of the regression: ns: p > 0.05; *: R2 < 0.1, p < 0.05; **: R2 ≥ 0.1, p < 0.01; ***: R2 ≥ 0.5, p < 0.001; ****: R2 ≥ 0.75, p < 0.0001. Points and regression lines per host are colour-coded as in Figs. 9-10. Shell length is plotted on a logarithmic scale.
Replication error in morphometric data
Shell lengths of a random selection of snails (n = 65) were measured in triplicate. Average distance from the mean of these three measurements was 0.03 mm for shell length. Error in landmark data was assessed based on 53 out of the 65 specimens (47 specimens of C. galea, six specimens of C. caribaea). Replication errors were slightly higher for C. galea than for C. caribaea. Error in digitizing landmarks was 3.6% in C. galea compared to 1.7% in C. caribaea. The total error, which included inconsistencies in parallax as well, was 11.4% in C. galea and 4.0% in C. caribaea. Replication error in C. curacaoensis sp. nov. was not calculated.
Sex could be determined for 609 specimens of C. galea and 115 specimens of C. caribaea. In C. galea, 74.4% were male. With 64.3%, the fraction of males in C. caribaea was lower (p = 0.030). Females were larger than males in both C. galea (χ2 = 133.9; p < 0.0001) and C. caribaea (F = 35.1; p < 0.0001). Both males and females were found on all but one host species (C. caribaea on Porites furcata Lamarck, 1816, on which only female specimens were found). Sex ratios within C. galea differed among snails associated with different host species (p = 0.0004; only including hosts with ≥ 5 specimens; Fig. 13a). After Bonferroni correction of p-values (n = 78) one pair remained significant: snails associated with Agaricia humilis Verrill, 1901 had a higher male to female ratio than snails associated with Colpophyllia natans (Houttuyn, 1772) (p = 0.032). Within C. caribaea, no variation in sex ratios between host species existed (p = 0.975; only including hosts with ≥ 5 specimens; Fig. 13b). Sex ratio of C. caribaea associated with alcyonaceans (66.0% male) and scleractinians (63.1% male) was similar as well (p = 0.845).
Within C. galea, mean length of females per host species correlated with the mean length of males (R2 = 0.707; p = 0.0002) (Fig. 14a). The same was observed in C. caribaea (R2 = 0.614; p = 0.011) (Fig. 14b).
The three Caribbean Coralliophila species clustered into three well-supported, separate clades on the phylogenetic tree (posterior probability (PP) = 0.997-1.000; Fig. 15). Phylogenetic trees constructed for both markers separately showed the same pattern (Online Supplementary Material 5). No monophyletic Caribbean cluster was found: Coralliophila curacaoensis sp. nov. is the sister species of a group consisting of both C. fontanangioyae Smriglio and Mariottini, 2000 and C. meyendorffii (Calcara, 1845) from the Eastern Atlantic and the Mediterranean; C. galea is sister of C. mira (Cotton and Godfrey, 1932) from the Indo-Pacific, and C. caribaea is sister of the Indo-Pacific Leptoconchus sp. Hence, the genus Coralliophila is also not monophyletic.
No host-associated genetic divergence was found within Coralliophila galea (p = 0.458 and p = 0.342 for 12S and COI, respectively). In C. caribaea on the other hand, a small divergence (mean uncorrected distance of 0.6% for 12S and 2.9% for COI) was found between snails associated with Alcyonacea and those associated with Scleractinia (p < 0.0001 for both 12S and COI). This genetic structuring within C. caribaea was also visible on the phylogenetic tree, as C. caribaea collected from Scleractinia cluster all on a single branch (with one exception), although the support value for this branch was low (PP = 0.803). There was no host-associated divergence within C. caribaea from scleractinian corals (p = 0.213 and p = 0.971 for 12S and COI, respectively) or alcyonacean hosts (p = 0.941 and p = 0.945 for 12S and COI, respectively).
The small divergence among C. caribaea individuals associated with either scleractinians or alcyonaceans was also observed in the haplotype networks constructed for both markers (Fig. 16). With one exception for COI, haplotypes were unique to snails associated with hosts from either Scleractinia or Alcyonacea. In addition, haplotypes of snails associated with alcyonaceans mostly (with one exception for both markers) clustered together on a single branch in the haplotype network. Snails associated with scleractinians formed (again, with one exception for both markers) the other branches in the networks. The two clusters were separated by two mutations for 12S and eight mutations for COI. No clear correlation between host species and haplotype was present within C. galea.
The ABGD analysis based on COI, using a more extensive dataset of coralliophiline snails, proposed nine different groupings of specimens into MOTUs, depending on the a priori boundary between intra- and interspecific divergence. A histogram of the frequencies of pairwise Kimura two-parameter (K2P) distances revealed a multimodal distribution, with the lowest minimum frequency around a K2P distance of 0.05 (Fig. 17). Using this value as a boundary between intra- and interspecific divergence, the coralliophiline dataset can be subdivided into 16 MOTUs. At this value, the specimens of C. galea, C. caribaea and C. curacaoensis sp. nov. are clustered into three different MOTUs, in agreement with the phylogenetic tree. The divergence within C. caribaea was confirmed to be intraspecific, not revealing any cryptic species, using this boundary value of 0.05.