Sharing the slope: depth partitioning of agariciid corals and associated Symbiodinium across shallow and mesophotic habitats (2-60 m) on a Caribbean reef

Abstract

Background: Scleractinian corals and their algal endosymbionts (genus Symbiodinium) exhibit distinct bathymetric distributions on coral reefs. Yet, few studies have assessed the evolutionary context of these ecological distributions by exploring the genetic diversity of closely related coral species and their associated Symbiodinium over large depth ranges. Here we assess the distribution and genetic diversity of five agariciid coral species (Agaricia humilis, A. agaricites, A. lamarcki, A. grahamae, and Helioseris cucullata) and their algal endosymbionts (Symbiodinium) across a large depth gradient (2-60 m) covering shallow to mesophotic depths on a Caribbean reef.

Results: The five agariciid species exhibited distinct depth distributions, and dominant Symbiodinium associations were found to be species-specific, with each of the agariciid species harbouring a distinct ITS2-DGGE profile (except for a shared profile between A. lamarcki and A. grahamae). Only A. lamarcki harboured different Symbiodinium types across its depth distribution (i.e. exhibited symbiont zonation). Phylogenetic analysis (atp6) of the coral hosts demonstrated a division of the Agaricia genus into two major lineages that correspond to their bathymetric distribution ("shallow": A. humilis / A. agaricites and "deep": A. lamarcki / A. grahamae), highlighting the role of depth-related factors in the diversification of these congeneric agariciid species. The divergence between "shallow" and "deep" host species was reflected in the relatedness of the associated Symbiodinium (with A. lamarcki and A. grahamae sharing an identical Symbiodinium profile, and A. humilis and A. agaricites harbouring a related ITS2 sequence in their Symbiodinium profiles), corroborating the notion that brooding corals and their Symbiodinium are engaged in coevolutionary processes.

Conclusions: Our findings support the hypothesis that the depth-related environmental gradient on reefs has played an important role in the diversification of the genus Agaricia and their associated Symbiodinium, resulting in a genetic segregation between coral host-symbiont communities at shallow and mesophotic depths.

Figure 1

Mesophotic habitat, sampling location, and agariciid abundances over depth. (a) Mesophotic Agaricia communities on the leeward shore of Curaçao at 50 m depth, (b) location of Curaçao (indicated by arrow) and study site Buoy Zero/One (indicated by red dot), and (c) distribution of Agaricia spp. and Helioseris cucullata over depth at the Buoy Zero/One study site.

Figure 2

Distribution of Symbiodinium ITS2 profiles across the five agariciid host species and depths. Each pie chart represents the sampled population of a host species at a certain depth. Asterisk (*) indicates a P4 profile that had an additional C1 band. Figure legend text.

Figure 3

Sequence network of Symbiodinium ITS2 types observed for each DGGE profile obtained from agariciid species. Ancestral Symbiodinium types C1 and C3 are represented by squares. Circle size is not representative of frequency. Dashed lines indicate sequence gaps (with gap size in number of base pairs). Colours group the different ITS2 sequences observed in each DGGE profile. The C3 type (indicated by black square) is found in all DGGE profiles except for P5.

Figure 4

Unrooted network of host atp6 mitochondrial haplotypes. Circle size indicates the relative frequency of each haplotype. Pie chart indicates for which agariciid species the haplotype was observed (Ahum = A. humilis (n = 16), Aaga = A. agaricites (n = 22), Alam = A. lamarcki (n = 27), Agra = A. grahamaei (n = 9), Asp. = Unidentified Agaricia species (n = 9), Hcuc = H. cucullata (n = 10)), and colour indicates the Symbiodinium profile that was hosted.

Figure 5

Phylogenetic trees of Agaricia spp. based on mitochondrial and nuclear markers. Phylogenetic trees (maximum parsimony) of Agaricia spp. based on (a) mitochondrial atp6 (this study), and (b) nuclear 28S and (c) mitochondrial COI (previous studies: Medina et al. 2006; Fukami et al. 2008; Shearer and Coffroth 2008; Barbeitos et al 2010). Pavona spp., Helioseris cucullata and Leptoseris sp. are used as outgroups. Bootstrap values are based on Bayesian (BAY), maximum parsimony (MP) and maximum likelihood (ML) respectively, with only probabilities over 50% shown. Depth range and number of specimens are mentioned for each species, coloured boxes indicates the Symbiodinium profile that was hosted. Sequences retrieved from GenBank are indicated by their Accession Number.

Figure 6

Conceptual diagram depicting niche partitioning of agariciid host-symbiont assemblages over depth. Relative abundances are extrapolated from average densities of host species and relative symbiont abundances across depths.

Figure 7

Photo of Agaricia community beyond the depth range of this study. Paper-thin Agaricia grahamae communities at 86 m depth on the leeward shore of Curaçao.