Ancestor–descendant relationships in evolution: origin of the extant pygmy right whale, Caperea marginata

Abstract

Ancestor–descendant relationships (ADRs), involving descent with modification, are the fundamental concept in evolution, but are usually difficult to recognize. We examined the cladistic relationship between the only reported fossil pygmy right whale, †Miocaperea pulchra, and its sole living relative, the enigmatic pygmy right whale Caperea marginata, the latter represented by both adult and juvenile specimens. †Miocaperea is phylogenetically bracketed between juvenile and adult Caperea marginata in morphologically based analyses, thus suggesting a possible ADR—the first so far identified within baleen whales (Cetacea: Mysticeti). The †Miocaperea–Caperea lineage may show long-term morphological stasis and, in turn, punctuated equilibrium.

1. Introduction

Descent with modification has been the essence of evolution since 1859 [1]. Ideally, an understanding of evolution would be based on ancestor–descendant relationships (ADRs). In reality, ADRs have usually been shown only for groups with abundant fossils in a dense stratigraphic record, including vertebrates [2], macro-invertebrates [3] and especially microfossils [4]. Since the development of cladistics [5] in the 1960s, emphasis has shifted from ADRs to phylogenetic sister-group relationships (SGRs). Indeed, fossils and ADRs have been discussed rather cursorily in the literature [6] as the debates of ancestry and cladistics in the 1970s and 1980s (e.g. pros: [7,8]; cons: [9,10]).

It is problematic to establish ADRs in vertebrate evolution, even with abundant fossils (e.g. [2]). Consider the filter-feeding baleen whales, the largest vertebrates [11], which clearly evolved from toothed stem Mysticeti [12–16]. The evolution of filter-feeding and large size involved dramatic structural and functional shifts in evolution. Species in the †Aetiocetidae were proposed to have teeth and baleen [12], potentially bridging the morphological gap between toothed Mysticeti and baleen-bearing Mysticeti. However, there is no close ADR between known toothed Mysticeti [12,13] and baleen-bearing Mysticeti [17,18]. Likewise, the oldest fossil mysticete, †Llanocetus denticrenatus [19] (late Eocene, Antarctica), is not clearly a direct descendant from known Archaeoceti and/or directly ancestral to any named mysticete. To establish ADRs, we must understand ancestral morphology either through fossils or through the proxy of juvenile morphology [20,21]. Here, we coded the morphology of an individual juvenile Caperea for use as a discrete operational taxonomic unit (OTU) in phylogenetic analyses. The adult and juvenile Caperea bracket the fossil †Miocaperea pulchra, producing a pattern that is consistent with ADRs, and that has implications for recognizing punctuated equilibrium in mysticete evolution.

2. Material and methods

We used matrices from articles on: (i) a new Caperea-like fossil, †Miocaperea pulchra (late Miocene, 7–8 Ma) [22], and (ii) a new phylogenetic interpretation of Caperea marginata [23]. We retained the original codings to optimize comparison with published results. The two matrices [22,23] cover diverse fossil and modern mysticetes, and produce different phylogenies: Bisconti [22] placed C. marginata close to right whales, Balaenidae, whereas Fordyce & Marx [23] placed C. marginata as an extant relict species in the family Cetotheriidae.

Initially, we examined 35 variably complete juvenile skeletons of Caperea marginata (electronic supplementary material, table S1). We considered that individual juveniles of slightly different ages/sizes could be merged into composite OTUs, or coded as separate OTUs, but we saw no clear objective way to discriminate age classes among the 35 juveniles and so chose one well-preserved specimen as a representative juvenile OTU. Specimen SAM M9079 (condylobasal length = 50.3 cm) was coded (photos of SAM M9079: supplementary files in [21]) and added to the original Bisconti matrix [22], which already contained adult C. marginata and †M. pulchra. Codings for †M. pulchra and the juvenile C. marginata were added to the Fordyce & Marx matrix [23], which already contained adult C. marginata. Details of TNT (v. 1.1) [24] settings and morphological scorings for phylogenetic analyses are in the electronic supplementary material and two separate TNT files.¹

3. Results

†Miocaperea pulchra consistently plots close to adult and/or juvenile Caperea marginata (figure 1, time-calibrated phylogenies). The Fordyce & Marx matrix [23] analysed with equal weights (figure 1a) shows †M. pulchra as the sister group to the juvenile C. marginata. With implied weights (figure 1b), †M. pulchra is basal to adult and juvenile C. marginata.

Figure 1.

Phylogenetic placement of †Miocaperea pulchra in relation to adult and juvenile Caperea marginata with stratigraphically calibrated trees. (a) Extracted phylogeny of Cetotheriidae, using data matrix of [23] with equally weighted setting. (b) Same as (a), but with implied weighting (k = 3). (c) Extracted phylogeny of Balaenoidea, using data matrix of [22] with equally weighted setting. (d) Same as (c), but with implied weighting (k = 3). For details of full phylogenies from each result, see electronic supplementary material. The number above each branch indicates how many characters change between the taxon and node. Dagger (†) indicates extinct species. juv, juvenile; Plei, Pleistocene; Holo, Holocene.

In the Bisconti matrix [22], †M. pulchra forms the sister taxon to adult C. marginata (figure 1c—equal weights; 1d—implied weights), while juvenile C. marginata represents the adjacent and most basal lineage in the Balaenoidea. Different synapomorphies drive the phylogenies from the two published matrices [22,23], placing †M. pulchra either basal to a clade comprising adult + juvenile C. marginata, or as the sister taxon to either the juvenile or the adult. Details of synapomorphies for each clade from two different matrices are in the electronic supplementary material.

4. Discussion

The ADR can be considered cladistically by using adult and juvenile specimens of a single species as separate OTUs, in expectation that the OTUs might then delimit what we term an ontogenetic clade. An ontogenetic clade would arise if the adult OTU plotted more crownward, and the conspecific juvenile plotted more basally. The phylogenetic position of a fossil, either within or outside the ontogenetic clade, provides a way to consider the ADR in evolution. In this case, one or other of the juvenile and adult Caperea marginata plot in a clade with †Miocaperea pulchra, or lie immediately adjacent. This ‘bracketing’ leads us to consider that †M. pulchra is ancestral to C. marginata. There are four possible evolutionary relationships for †Miocaperea and Caperea (figure 2):

• (1) the most recent common ancestor of †Miocaperea + Caperea is unknown (figure 2a) (SGR),

• (2) †Miocaperea is the ancestor of Caperea (figure 2b,c), with †Miocaperea–Caperea representing an anagenetic lineage (figure 2b) (ADR),

• (3) as for (2) but with Caperea split (by cladogenesis) from †Miocaperea (figure 2c) (ADR),

• (4) †Miocaperea split from a long Caperea lineage, subsequently going extinct (figure 2d), making Caperea ancestral to †Miocaperea (ADR).

Figure 2.

Four possible evolutionary scenarios of †Miocaperea–Caperea relationships with a stratigraphically calibrated diagram. (a) An unknown common ancestor that remains undetected. (b) †Miocaperea–Caperea forms an anagenetic lineage. (c) Caperea split from †Miocaperea. (d) †Miocaperea split from Caperea. Black square denotes the ancestor of living Caperea. Plei, Pleistocene; Holo, Holocene.

Especially for fossil vertebrates, there is a low probability of recovering ancestors from the fossil record, and thus a low likelihood of recognizing ADRs [25]. The improbability of finding ancestors would allow scenario 1 (figure 2a) to represent †Miocaperea–Caperea evolution. Results do not preclude a SGR for †Miocaperea–Caperea, with an unknown ancestor at the branching node, rather than the ADR of scenario 1.

Consider two aspects of ADRs for †Miocaperea–Caperea. First, when †Miocaperea is bracketed cladistically between juvenile and adult Caperea, †Miocaperea could be ancestral to Caperea, not merely in a SGR. Whether anagenesis (figure 2b) or cladogenesis (2c) is involved remains uncertain. Scenario 4, with C. marginata ancestral to †M. pulchra (figure 2d), is unlikely, since adult C. marginata is derived relative to †M. pulchra in some features (e.g. anterior elongation of the pars cochlearis, and other character states in electronic supplementary material).

Second, C. marginata shows paedomorphic neoteny, which reduces the rate of morphological development during early ontogeny [26]. Given the link between ontogeny and phylogeny [20], paedomorphic neoteny could scale into long-term (7–8 Ma) morphological stasis in the †Miocaperea–Caperea lineage. Stasis is, in turn, a key aspect of punctuated equilibrium [27,28]. Our results, and the minor morphological differences noted by Bisconti between †Miocaperea and Caperea, are consistent with minimal long-term morphological change in the lineage, allowing punctuated equilibrium—geologically rapid origination followed by a long interval of relative stasis [27,28].

The enigmatic Caperea marginata has been the sole species in Caperea and in the Neobalaenidae (or Cetotheriidae: Neobalaeninae) since its recognition in the mid-1800s, with no obvious fossil relatives reported until Bisconti described the similar Miocene †M. pulchra in 2012. Possible morphological stasis (punctuated equilibrium) in the single lineage of †Miocaperea–Caperea since the late Miocene (7–8 Ma) could explain the low diversity of neobalaenines relative to rorquals (balaenopterids). Strongly pelagic habits could explain the poor fossil record of the Caperea lineage, precluding estimates of former neobalaenine diversity and stratigraphic ranges (e.g. [29–31]). The morphological similarity between †Miocaperea and Caperea raises a taxonomic issue—whether †Miocaperea pulchra is sufficiently different morphologically to be generically distinct from Caperea. For now, we follow the name †Miocaperea pulchra. Interestingly, if †Miocaperea were reconsidered as in the genus Caperea, this long-term morphological stasis between †Miocaperea–Caperea may represent stabilomorphism [32]. Of note, here we consider that living Caperea may be descended from the fossil species †Miocaperea, but Caperea should not be considered as living fossil [33].

Recent debate on Caperea phylogeny [23,34,35] reflects different character selections and interpretations of homology, as results here also indicate (see electronic supplementary material). This highlights the issues of character selection for phylogeny in the first place [36] and recognition of homology [37], as well as the distinctive evolutionary history of Caperea [26]. Perhaps the most persuasive instance of ADRs among marine mammals involves sirenians in the lineage leading to Steller's seacow, Hydrodamalis, as inferred from morphological change in fossils [38]. We show now that phylogenetic approaches might help to recognize ADRs in the †Miocaperea–Caperea lineage, providing a new look at neobalaenine evolution in spite of the sparse fossil record. By exploring broad patterns and processes in baleen whale evolution [21,26], including ADRs (this study), and describing more species (e.g. [12,13,17–19,22,34,35]), a better picture of mysticete evolution—involving the largest animals—will arise.

Data accessibility

Data are available at doi:10.5061/dryad.21q24.

Author contributions

C.-H.T. designed the research and conducted the phylogenetic analyses. C.-H.T. and R.E.F. collected data, and wrote the paper.

Funding statement

C.-H.T. was supported by a University of Otago Doctoral Scholarship.

Competing interests

Authors have declared that no competing interests exist.