Multiple evolutionary transitions of reproductive strategies in a phylum of aquatic colonial invertebrates

Abstract

Parental care is considered crucial for the enhanced survival of offspring and evolutionary success of many metazoan groups. Most bryozoans incubate their young in brood chambers or intracoelomically. Based on the drastic morphological differences in incubation chambers across members of the order Cheilostomatida (class Gymnolaemata), multiple origins of incubation were predicted in this group. This hypothesis was tested by constructing a molecular phylogeny based on mitogenome data and nuclear rRNA genes 18S and 28S with the most complete sampling of taxa with various incubation devices to date. Ancestral character estimation suggested that distinct types of brood chambers evolved at least 10 times in Cheilostomatida. In Eucratea loricata and Aetea spp. brooding evolved unambiguously from a zygote-spawning ancestral state, as it probably did in Tendra zostericola, Neocheilostomata, and 'Carbasea' indivisa. In two further instances, brooders with different incubation chamber types, skeletal and non-skeletal, formed clades (Scruparia spp., Leiosalpinx australis) and (Catenicula corbulifera (Steginoporella spp. (Labioporella spp., Thalamoporella californica))), each also probably evolved from a zygote-spawning ancestral state. The modular nature of bryozoans probably contributed to the evolution of such a diverse array of embryonic incubation chambers, which included complex constructions made of polymorphic heterozooids, and maternal zooidal invaginations and outgrowths.

Keywords: Bryozoa; brooding; endotrophy; larva; planktotrophy; viviparity.

Figure 1.

Reproductive structures of cheilostome bryozoans. (a) Membranipora membranacea with two expanded lophophores having an intertentacular organ (arrows and dotted lines); ripe zygotes shown by asterisks (photo by A. Gruhl, North Sea). (b) Autozooid of Aetea sp. with embryo inside external membranous sac (arrow) (NHMUK 1947.4.22.341). (c) Steginoporella sp. showing three fertile zooids with embryos (arrows) inside internal brood sacs; young oocyte shown by asterisk (photo by T. Schwaha, Great Barrier Reef, Australia). (d) Tendra zostericola showing two zooids with acanthostegal brood chambers (arrows) (inset: upper view of tentacle crown showing round mouth centrally and intertentacular organ to the left, dotted line) (NHMW 2022.VII.1). (e) Catenicula corbulifera showing fertile autozooid with synoecium (arrow) of six kenozooids (numbered 1–6) with cuticular joints (NHMUK 2023.6.22.6) (photo by Silviu Martha). (f) Scruparia ambigua showing fertile autozooid with bivalved brood chamber (arrow) made of two kenozooids (eastern Atlantic, northern Spain). (g) Thalamoporella crenulata showing fertile autozooid with bipartite brood chamber (arrow) containing three embryos (photo by T. Schwaha, Great Barrier Reef, Australia). (h) Puellina egretta showing three fertile zooids with ovicells (arrows) (Red Sea). NHMUK, Natural History Museum London; NHMW, Naturhistorisches Museum Wien.

Figure 2.

Bayesian phylogenetic analysis of the mixed concatenated alignment consisting of three partitions: (i) 13 mitochondrial protein-coding genes as amino acids, (ii) mitochondrial ribosomal RNA genes 12S + 16S, (iii) nuclear ribosomal RNA genes 18S + 28S. The analysis was performed in MrBayes5D v. 3.2.6 under the GTR + G model of nucleotide evolution (nucleotides) and the MTZOA + G model (amino acids). The analysis was run for 2.4 million generations; 1.5 million generations were discarded as burn-in. Posterior probabilities are given above the branches. Maximum-likelihood bootstrap support values generated using RAxML HPC-PTHREADS-SSE3 v. 8.2.12 over 100 replicates are given below the branches. The branch length-scale bar indicates number of substitutions per site. Higher level classification is given at the right-hand side. Cheilostomate suborders are indicated with different colours. Malacostegan-grade taxa are in light blue font. Whenever multiples of the same species/genus sp. are included, countries of origin or specimen codes are given.

Figure 3.

Ancestral character estimation of reproductive modes performed in phytools using the ace function and the equal-rates (ER) model implemented in the R package ape. The empirical Bayesian posterior probabilities (EBPP) which indicate the probability of occurrence of character states is given as pie charts at nodes. For EBPP values corresponding to the numbered nodes, see electronic supplementary material, table S6. Asterisks and associated dashed lines indicate the distribution of distinct brood chamber types.