From the male peacock's tail plumage to the floral displays of flowering plants, traits related to sexual reproduction are often complex and exaggerated. Why has sexual reproduction become so complicated? Why have such exaggerated sexual traits evolved? Early work posited a connection between multicellularity and sexual traits such as anisogamy (i.e., the evolution of small sperm and large eggs). Anisogamy then drives the evolution of other forms of sexual dimorphism. Yet, the relationship between multicellularity and the evolution of sexual traits has not been empirically tested. Given their extensive variation in both multicellular complexity and sexual systems, the volvocine green algae offer a tractable system for understanding the interrelationship of multicellular complexity and sex. Here we show that species with greater multicellular complexity have a significantly larger number of derived sexual traits, including anisogamy, internal fertilization, and secondary sexual dimorphism. Our results demonstrate that anisogamy repeatedly evolved from isogamous multicellular ancestors and that anisogamous species are larger and produce larger zygotes than isogamous species. In the volvocine algae, the evolution of multicellularity likely drives the evolution of anisogamy, and anisogamy subsequently drives secondary sexual dimorphism. Multicellularity may set the stage for the overall diversity of sexual complexity throughout the Tree of Life.
combined_chloroplast_collection
Systematic sample of Bayesian trees. Every 1,000th post-burnin tree from the four runs was included, for a total of 600 trees.
Volvocine tree, ultrametric
Volvocine phylogenetic tree, created from the phylogram using a penalized likelihood function.
combined_chloroplast.con.clado.nex
Volvocine tree, phylogram
Bayesian phylogram of volvocine and outgroup taxa.
combined_chloroplast.con.tre.nex
Chloroplast data matrix
Data matrix for volvocine chloroplast phylogenetic tree.
combined_chloroplast.fasta.nex
Dataset 1
Genbank accession numbers of chloroplast genes for ingroup and outgroup taxa.
Dataset 2
Continuous multicellularity metrics and discrete character states for sexual traits and associated references for volvocine taxa. For heterothallic species, only one strain of a mating pair has been given.
Table A1
PCR primers used in this study. *indicates a reverse primer
Table A2
The best partition and nucleotide substitution scheme as determined by PartitionFinder version 2.1.1 (Lanfear et al. 2016). “Position” refers to codon position.
Table A3
Model fit and parameter estimates for analyses of discrete traits. For phylogenetic signal, the D value was calculated (D < 0 indicates stronger phylogenetic signal than a Brownian motion model; D = 0 indicates Brownian motion; D = 1 indicates random, non-phylogenetic distribution of states among taxa). Significant P-values associated with a given model indicate that the model is rejected. For discrete traits, the first state listed in the column header is state 0, and subsequent states are state 1 or state 2. The value q is the transition rate between the two states (ancestral state on the left, derived state on the right). Bayes factors (BF), taking phylogenetic uncertainty into account, were estimated based on twice the difference between the highest harmonic mean log likelihood for each model (ER, equal rates; SYM, symmetric rates; ARD, all rates different), calculated from nine independent MCMC runs for each model. Interpretation of Bayes factors is as follows: 0 to 2 barely worth mentioning, 2 to 6 positive, 6 to 10 strong, >10 very strong (Kass and Raftery 1995).
Table A4
Model fit and parameter estimation for analyses of continuous traits. Phylogenetic signal was assessed using Blomberg’s K (2003) and Pagel’s λ (Pagel 1999). Statistical significance for these indices was based on comparison to a non-phylogenetic model (K = 0, λ = 0), with a significant P-value indicating rejection of the non-phylogenetic model.
Table A5
Correlations between internal fertilization and anisogamy and oogamy using Pagel’s test (1994)
Table 1
Phylogenetic t-test for sexual traits, using three continuous metrics of multicellular complexity (rounds of cell division, percent somatic cells, natural logarithm-transformed body length). A phylogenetic t-test comparing natural logarithm-transformed zygote diameter of isogamous and anisogamous algae is included. When multiple metrics were used, p-values were adjusted for multiple comparisons following Benjamini and Hochberg (1995).
