Shedding light: a phylotranscriptomic perspective illuminates the origin of photosymbiosis in marine bivalves

Abstract

Background: Photosymbiotic associations between metazoan hosts and photosynthetic dinoflagellates are crucial to the trophic and structural integrity of many marine ecosystems, including coral reefs. Although extensive efforts have been devoted to study the short-term ecological interactions between coral hosts and their symbionts, long-term evolutionary dynamics of photosymbiosis in many marine animals are not well understood. Within Bivalvia, the second largest class of mollusks, obligate photosymbiosis is found in two marine lineages: the giant clams (subfamily Tridacninae) and the heart cockles (subfamily Fraginae), both in the family Cardiidae. Morphologically, giant clams show relatively conservative shell forms whereas photosymbiotic fragines exhibit a diverse suite of anatomical adaptations including flattened shells, leafy mantle extensions, and lens-like microstructural structures. To date, the phylogenetic relationships between these two subfamilies remain poorly resolved, and it is unclear whether photosymbiosis in cardiids originated once or twice.

Results: In this study, we establish a backbone phylogeny for Cardiidae utilizing RNASeq-based transcriptomic data from Tridacninae, Fraginae and other cardiids. A variety of phylogenomic approaches were used to infer the relationship between the two groups. Our analyses found conflicting gene signals and potential rapid divergence among the lineages. Overall, results support a sister group relationship between Tridacninae and Fraginae, which diverged during the Cretaceous. Although a sister group relationship is recovered, ancestral state reconstruction using maximum likelihood-based methods reveals two independent origins of photosymbiosis, one at the base of Tridacninae and the other within a symbiotic Fraginae clade.

Conclusions: The newly revealed common ancestry between Tridacninae and Fraginae brings a possibility that certain genetic, metabolic, and/or anatomical exaptations existed in their last common ancestor, which promoted both lineages to independently establish photosymbiosis, possibly in response to the modern expansion of reef habitats.

Keywords: Fraginae; Photosymbiosis; Reef habitat; Symbiodiniaceae; Tridacinae.

Fig. 1

Morphological and ecological comparisons among Tridacninae (Tridacna squamosa) and Fraginae (Fragum fragum, Corculum cardissa) species. a. Lateral shell views of the three species and diagrams showing their typical positions in natural habitats. Yellow rectangles represent sediment. Note that F. fragum and C. cardissa both show different degrees of posterior shell compression. b–d. Photos of the three species in their natural habitat. F. fragum (c) was taken out of the sediment when photo was taken. Photo credits: Jingchun Li and Jeff Whitlock (the Online Zoo). Images were processed in Affinity Designer 1.8.4 (Serif Ltd.)

Fig. 2

Gene occupancy diagram showing the 25 matrices analyzed in this study. Top: Main matrices of two minimum taxon occupancy thresholds. Orthogoups in Matrix 1 and 2 are shared by at least 50 and 75% of all taxa. Bottom: Matrix 1 was divided into 22 sub matrices based on gene evolution rates from slow (A) to fast (V). Each matrix containing 52 orthogroups, except for the last matrix V, which contains 16 orthogroups

Fig. 3

a-b. The two best supported topologies obtained from the analyses of the two main matrices and 22 submatrices. Cardiid subfamilies are indicated by different colors. Supporting matrices and corresponding analytical methods are listed under each topology. c. Phylogenetic results based on maximum likelihood analysis (PhyML-PCMA) of Matrix 1 and Bayesian analysis (PhyloBayes) of Matrix 2. This is also the topology supported by the most analyses. Node labels represent bootstrap supports / posterior probabilities of each subfamily and the backbone. Photosymbiotic clades are shaded in grey. Shell position of Fraginae and Tridacninae species in their natural habitats is shown in the two diagrams

Fig. 4

Fossil calibrated phylogeny and ancestral state reconstruction based on the most supported topology. Letters (a–d) at nodes indicate calibration points. Blue bars at nodes represent standard deviation of age estimation. Red and black bars at nodes represent probabilities of the common ancestor being photosymbiotic (red) or non-photosymbiotic (black). The grey shading in the background corresponds to number of global reef sites through time (following [28])