Professional Documents
Culture Documents
Cetacean Feeding Evolution and Ecology PDF
Cetacean Feeding Evolution and Ecology PDF
SYMPOSIUM
Synopsis The origin of baleen and filter feeding in mysticete cetaceans occurred sometime between approximately 34
and 24 million years ago and represents a major macroevolutionary shift in cetacean morphology (teeth to baleen) and
ecology (raptorial to filter feeding). We explore this dramatic change in feeding strategy by employing a diversity of tools
and approaches: morphology, molecules, development, and stable isotopes from the geological record. Adaptations for
raptorial feeding in extinct toothed mysticetes provide the phylogenetic context for evaluating morphological apomor-
phies preserved in the skeletons of stem and crown edentulous mysticetes. In this light, the presence of novel vascular
structures on the palates of certain Oligocene toothed mysticetes is interpreted as the earliest evidence of baleen and
points to an intermediate condition between an ancestral condition with teeth only and a derived condition with baleen
only. Supporting this step-wise evolutionary hypothesis, evidence from stable isotopes show how changes in dental
chemistry in early toothed mysticetes tracked the changes in diet and environment. Recent discoveries also demonstrate
how this transition was made possible by radical changes in cranial ontogeny. In addition, genetic mutations and the
possession of dental pseudogenes in extant baleen whales support a toothed ancestry for mysticetes. Molecular and
morphological data also document the dramatic developmental shifts that take place in extant fetal baleen whales, in
skull development, resorption of a fetal dentition and growth of baleen. The mechanisms involved in this complex
evolutionary transition that entails multiple, integrated aspects of anatomy and ecology are only beginning to be under-
stood, and future work will further clarify the processes underlying this macroevolutionary pattern.
ß The Author 2016. Published by Oxford University Press on behalf of the Society for Integrative and Comparative Biology.
All rights reserved. For permissions please email: journals.permissions@oup.com.
1272 A. Berta et al.
Lunging
Skimming
Suctioning
Feeding pits
natural sieve to trap prey items during bulk feeding All extant mysticetes share four fundamental func-
(Pivorunas 1979; Werth 2001, 2004, 2013). Baleen is tional components related to bulk feeding: a large
composed of keratin (Fudge et al. 2009; Szewciw et head and mouth, rostral bones that are rather loosely
al. 2010) and consists of a serially arranged row of articulated with the braincase and with each other,
keratinous laminae and bristles that grow as a pair of an elastic mandibular symphysis, and baleen. All of
subparallel longitudinal racks from the right and left those features help to either filter food from water
lateral margins of the palate (Williamson 1973; (e.g., baleen), or to expand the oral cavity to facili-
Werth 2001; Young et al. 2015). Individual baleen tate an intake of large volumes of water.
laminae are roughly triangular in shape and are di- Morphological variations in these general mysticete
visible into main plates, minor plates, and individual features as exhibited by living baleen whales include
bristles. The main and minor plates consist of a cen- further enlargements of the oral cavities, which can
tral, medullary layer of keratin tubules enmeshed in be viewed as adaptations for different strategies of
an amorphous keratin matrix. On either side of the bulk filter feeding (Fig. 1), including continuous
medullary layer is a cortical layer of amorphous ker- ram or skim feeding in balaenids (bowhead and
atin. Each rack of baleen is deeply rooted within right whales) and presumably neobalaenids (pygmy
vascularized gingival tissues (Utrecht 1965). Baleen right whale), intermittent ram or engulfment/lunge
consists proximally of living tissue that receives feeding in balaenopterids (rorquals and humpback
blood from the superior alveolar arteries (Ekdale et whales), and suction feeding in eschrichtiids (gray
al. 2015), and distally of the dead and cornified whales) (see Supplementary Material).
tissue that extends into the oral cavity. As baleen One outstanding question in the acquisition of prey
grows in the mouth, the lingual border of the by mysticetes is how baleen whales detect and locate
plates is abraded and frayed to expose the bristles their food. Our understanding of prey detection in
(Werth et al. 2016). The free bristles on one odontocete cetaceans suggests that biosonar (‘‘echolo-
lamina overlap with the bristles of adjacent laminae cation’’) employing high frequency and ultrasonic
to form a dense, brush-like mat—the baleen filter. sounds is a primary mechanism (e.g., Ketten 2004;
Filter feeding in mysticete cetaceans 1273
Nachtigall et al. 2007). However, mysticetes are likely comparative anatomical studies of fetal specimens
sensitive to low and perhaps infrasonic frequencies (e.g., Ridewood 1923; Dissel-Scherft and Vervoort
(Houser et al. 2001; Erbe 2002; Parks et al. 2007) and 1954; Karlsen 1962; Tomilin 1967; Slijper 1976;
lack echolocation capabilities. Likewise, the eyes of Fudge et al. 2009). Although neonate, juvenile, and
extant cetaceans are not particularly adapted for en- adult mysticetes are edentulous, tooth buds develop
hanced visual acuity (Mass and Supin 1997; Buono et in both the upper and lower jaws in utero. However,
al. 2012; Rodrigues et al. 2014, 2015); however, en- the teeth are resorbed completely and the baleen ap-
hanced visual acuity may have been characteristic of paratus begins to develop before birth. In the minke
some early diverging extinct mysticetes as hypothesized whale (Balaenoptera acutorostrata), the developing
by Fitzgerald (2006) and quantified by Marx (2011). By teeth pass through three of the typical mammalian
contrast with extinct mysticetes, molecular evidence stages (bud, cap, bell), but they never erupt and
Given that mysticetes possess tooth buds at some (McKnight and Fisher 2009). Genes in the enamel
ontogenetic stages, it is no surprise that they retain matrix proteins domain (AMBN, AMEL) acquired
functional tooth-related genes from their toothed an- inactivating mutations in different exons in all lineages
cestors. Several molecular studies investigated puta- of toothless amniotes, including baleen whales
tive enamel and tooth-related genes in mysticetes (Meredith et al. 2013), but the mutations accumulated
(Deméré et al. 2008; McKnight and Fisher 2009; in different exons among different mysticete species.
Meredith et al. 2009, 2011, 2013, Springer et al. Another gene of the same domain, ENAM, was found
2016b), thereby providing a better understanding of to have become a pseudogene in crown mysticetes, even
the functions and importance of these gene regions if the inactivating mutations are not shared among the
(Fig. 3). A review of those data reveals that loss of genera that were analyzed. Another putative enamel-
enamel specifically and teeth in general was a step- specific gene (C4orf26; Meredith et al. 2011) was
wise process during mysticete evolution that was also found to have a complex evolutionary history—
accomplished via different pathways in different mys- balaenids retain most of the sequence with inactivating
ticete lineages. mutations, while most but not all balaenopteroid spe-
Certain genes associated with teeth/tooth formation cies lack the gene completely (Springer et al. 2016b).
in mammals (WDR72, SLC24A4, FAM83H) have been The blue whale (Balaenoptera musculus) possesses the
found to be mostly conserved in at least three mysticete gene, but it is inactivated by mutations different from
species, implying that they might serve other important those of Balaenidae. The most parsimonious hypothesis
functions in development (Springer et al. 2016b). to explain the various genetic patterns suggests that the
Similarly, the DPP region of the dentin-related DSPP different mutations are a result of ancestral polymor-
(dentin sialophosphoprotein) gene has been found to phisms and that some of them have been lost indepen-
be conserved at least in the bowhead whale, while it has dently in the extant species (Meredith et al. 2009). In
accumulated mutations in toothless xenarthrans contrast, MMP20 (matrix metalloproteinase) that is
Filter feeding in mysticete cetaceans 1275
involved in tooth mineralization and also has pleiotro- understand the tooth to baleen transition in
pic effects on other organs, has the same inactivating mysticetes.
mutation in all mysticetes, suggesting that the gene lost
functionality only once in the common ancestor of
Integrating neontological and paleonto-
Mysticeti. It should also be noted that the occurrence
of this shared mutation implies that enamel was lost
logical data
once in the common ancestor of crown Mysticeti The fossil record of mysticetes provides direct ana-
rather than independently in different lineages. tomical evidence for the tooth to baleen transition
Most of the current genetic research into the loss suggested by developmental patterns in extant fetal
of teeth in mysticetes has focused on genes related to specimens. This record spans approximately 35 mil-
tooth development, but the genetic basis for devel- lion years of Earth history from near the Eocene–
opment of the baleen apparatus itself is unclear. Oligocene boundary to the Recent (Fig. 4), and
Given the tissue’s keratinous composition, baleen mysticete fossils are well documented across a
genes are likely to have originated from hair genes nearly worldwide distribution. The Oligocene por-
that are expressed in the mouth of these mammals tion of this record includes specimens of toothed
that have lost the body hair cover (Pivorunas 1979). mysticetes from the North and South Pacific
It has been hypothesized that an acceleration in ex- (Fitzgerald 2006, 2010; Deméré and Berta 2008;
pression of genes of the WNT family during devel- Marx et al. 2015; Fordyce and Marx 2016) and
opment can explain both the hair loss and the stem edentulous mysticetes from the Southeastern
growth of baleen in mysticetes (Park et al. 2015). United States, Japan, and New Zealand (Sanders
Future investigations into the genetic basis of and Barnes 2002; Okazaki 2012; Boessenecker and
baleen development will be necessary to more fully Fordyce 2015a, 2015b, 2015c; Tsai and Fordyce,
1276 A. Berta et al.
2 Odontoceti
1 4 Mammalodontidae†
3 6 Aetiocetidae†
5 8 Eomysticetidae†
10
11 Balaenidae
9
stem Thalassotherii†
13 Cetotheriidae†
12
15 Eschrichtiidae
14
16 Balaenopteridae
Eocene Oligocene Miocene Plio.
40 30 20 10 0
(Millions of Years) Pleistocene Holocene
Feeding Regimes
suction-aided raptorial feeding baleen-aided raptorial feeding bulk lter feeding
skim lter feeding lateral suction lter feeding engulfment/ lunge lter feeding
Fig. 4 Cladogram of Cetacea with associated feeding strategies in extinct and extant mysticetes. Relationships based on a strict consensus of
cladograms presented by Boessenecker and Fordyce (2015b). Stratigraphic ranges were downloaded from the Paleobiology Database (paleo-
biodb.org) on June 9, 2016 by searching for geologic strata through the taxonomic name search form for members of each terminal taxon named
on the cladogram. Illustrations of whales by C. Buell. (1) Cetacea: 28–59% olfactory genes non-functional. (2) Odontoceti: 74–100% olfactory
receptor genes non-functional; biosonar (‘‘echolocation’’). (3) Mysticeti: broad rostrum; thin lateral margins of maxilla; small, wide teeth with
denticles; diastemata in upper jaw. (4) Mammalodontidae: enlarged forward facing eyes (enhanced visual acuity?). (5)
Aetiocetidae þ Chaeomysticeti: small lateral nutrient foramina (presence of baleen-like structure); mandibular symphysis unfused. (6)
Aetiocetidae: enlarged forward facing eyes (enhanced visual acuity?). (7) Chaeomysticeti: low 13C in bones (feeding at low trophic level); reduction
of postnatal dentition; long, flattened rostrum; premaxilla/maxilla kinesis; elaborate lateral nutrient foramina; reduced function of tooth and
enamel genes. (8) Eomysticetidae: vestigial anterior dentition. (9) Crown Mysticeti: definitive presence of baleen/postnatal teeth absent; anterior
extension of occipital shield; lateral bowing of dentaries; reduced function of tooth and enamel genes. (10) Balaenoidea: arching of rostrum; long
baleen; subrostral gap. (11) Balaenidae: extreme arching of rostrum. (12) Thalassotherii: anteriorly convergent baleen racks; indistinct/concave
glenoid fossa; loss of C4orf tooth genes (?) (13) Cetotheriidae: reduced gaping of mouth; semi-synovial jaw joint. (14) Balenopteroidea: loss of C4orf tooth
genes. (15) Eschrichtiidae: slight arching of rostrum; short, coarse baleen; semi-synovial jaw joint; few, short throat grooves. (16) Balaenopteridae:
synovial jaw joint lost; tongue reduced and flaccid; ventral throat pouch; mandibular sense organ; elongation and multiplication of throat grooves.
2015). Given the presence of functional adult teeth The earliest known stem toothed mysticete is
in toothed mysticetes, the genomes of these early Llanocetus denticrenatus from the latest Eocene
mysticetes certainly retained functional tooth and (34 Ma) of Antarctica. However, description of
enamel genes. only a partial dentary and dentition of this
Filter feeding in mysticete cetaceans 1277
important fossil are available, while the skull remains Northern Hemisphere, which represent a monophy-
undescribed. A similar problem involves a series of letic lineage of toothed mysticetes known from
undescribed archaeocete-like toothed mysticetes from Oligocene rocks (31–24 Ma) on both sides of the
the Oligocene (30–26 Ma) of South Carolina, United North Pacific (node 6, Fig. 4) and often described as
States that are based on well-preserved skulls, mandi- bridging the morphological gap between stem
bles, and postcrania. As a consequence, the earliest toothed and stem baleen-bearing mysticetes (e.g.,
well-described stem toothed mysticetes are represented Deméré and Berta 2008). Aetiocetids include five
by fossils from the Oligocene (28–24 Ma) of genera and numerous species: Aetiocetus,
Australia and New Zealand—Janjucetus hunderi, Chonecetus, Ashorocetus, Fucaia, and Morawanocetus,
Mammalodon colliveri, and Mammalodon hakataramea all of which have skulls with closed rostral sutures
(Fitzgerald 2006, 2010, 2011; Fordyce and Marx 2016). between maxilla and premaxilla, maxilla and vomer,
mammalodontids). These contrasting interpretations more rigid rostrum in Waharoa ruwhenua than in
of aetiocetid phylogeny and function underscore the crown mysticetes. The palatal morphology of
important role that this transitional group plays in Waharoa is also intermediate between toothed and
reconstructing early mysticete evolution and how the crown mysticetes. Unlike basal toothed mysticetes
anatomical features of purely suction-aided raptorial (e.g., Janjucetus and Mammalodon) but similar to
feeders could be co-opted in a step-wise fashion for aetiocetids (e.g., species of Aetiocetus), the palate of
bulk filter feeding in later diverging edentulous Waharoa has lateral palatal vascular features; how-
mysticetes. ever, the palatal foramina and sulci are larger and
Whether or not baleen was present, aetiocetid teeth more numerous than in aetiocetids. Waharoa differs
display a clear and interesting dichotomy between from extant mysticetes in having lateral nutrient fo-
species of Morowanocetus and Fucaia with distinctly ramina and sulci confined to the posterior two-thirds
were feeding at the same trophic level as toothed divergence of chaeomysticetes. That hypothesis is
whales and likely were not filter feeding or feeding based on fossils of Horopeta umarere from the late
low in the food chain as in extant mysticetes. In Oligocene (27–25 Ma) of New Zealand (Tsai and
contrast 13C values for eomysticetids and stem balae- Fordyce 2015). Horopeta is known from a partial skull
nopteroids are lower and significantly different from that is missing anatomically important parts of the basi-
those of toothed mammalodontids or odontocetes, cranium, squamosal, and rostrum. The mandibles are
suggesting that the earliest edentulous mysticetes fragmentary and bioeroded, but preserve just enough of
may have specialized on a diet of zooplankton sim- the horizontal rami and coronoid processes to distin-
ilar to extant mysticetes (Supplementary Fig. S4). guish them from eomysticetids. Tsai and Fordyce
Although aetiocetids were not included within the (2015) used only equivocal mandibular features (a
isotopic analyses, the results of those analyses pro- slightly laterally bowed horizontal ramus and a poster-
eschrichtiids (node 15, Fig. 4) and balaenopterids (node laminae; extremely telescoped supraoccipital, and re-
16, Fig. 4). A third group is more problematic, but in duced coronoid process) related to the continuous
general consists of a group of stem thalassotherians that, ram feeding behavior of Caperea. In addition, the
depending on the study examined, represents either an hypothesis suggests that continuous ram feeding
unnamed monophyletic group (Boessenecker and evolved independently within a single lineage of the
Fordyce 2015a, 2015b, 2015c) or a paraphyletic group cetotheriid clade, a clade that only contains suction
(Marx and Fordyce 2015; Marx et al. 2016). In either feeding mysticetes. Clearly, neobalaenids are an ex-
case, this problematic assemblage referred to here as tremely autapomorphic group, but few independent
stem thalassotherians lack the synapomorphies of research groups have been able to replicate this
cetotheriids and balaenopteroids. unique phylogenetic placement of Caperea as a
As expected in such a taxonomically diverse clade, cetotheriid. One exception is the study by Gol’din
high percentage of pachyosteosclerotic bone. A func- correlated with an increased reliance on the neck
tional analysis of C. riabinini suggests that this Eastern musculature.
Paratethyan cetotheriid was primarily a benthic suction Balaenopteridae is the most specious group of crown
feeder intermediate between continuous suction fee- thalassotherians with 8 or 9 living members and at least
ders such as Herpetocetus and intermittent suction fee- 13 described extinct taxa. The balaenopterid fossil
ders such as Eschrichtus (Gol’din and Startsev 2014). record extends into the late Miocene (12–10 Ma)
Extant thalassotherians form the clade and appears to contain several recognizable lineages.
Balaenopteroidea (node 14, Fig. 4) that includes A basal lineage consisting of Miocene and Pliocene
living and fossil members of Eschrichtiidae (gray species from Italy and California (Plesiobalaenoptera
whales; node 15, Fig. 4) and Balaenopteridae (rorquals; quarantellii, ‘‘Balaenoptera’’ ryani, Cetotheriophanes
node 16, Fig. 4). All extant balaenopteroids lack the capellinii) is characterized by skulls with elongate inter-
Dissel-Scherft MCV, Vervoort W. 1954. Development of the tooth development and degeneration in fetuses of minke
teeth in fetal Balaenoptera physalus (L.) (Cetacea, whale, Balaenoptera acutorostrata. J Vet Med Sci 61:227–32.
Mystacoceti). Proc K Ned Akad Wet Ser C 57:196–210. Ishishima H, Sato ET, Sagayama, Kimura M. 2006. The oldest
Drake SE, Crish SD, George JC, Stimmelmayr R, Thewissen record of Eschrichtiidae (Cetacea: Mysticeti) from the late
JGM. 2015. Sensory hairs in the bowhead whale, Balaena Pliocene, Hokkaido, Japan. J Paleontol 80:367–79.
mysticetus (Cetacea, Mammalia). Anat Rec 298:1327–35. Ivashin MV, Mikhalev YA. 1978. To the problem of the pre-
Ekdale EG, Deméré TA, Berta A. 2015. Vascularization of the natal growth of minke whales Balaenoptera acutorostrata of
gray whale palate (Cetacea, Mysticeti, Eschrichtius robustus): the Southern Hemisphere and of the biology of their re-
soft tissue evidence for an alveolar source of blood to production. Rep Intern Whaling Comm 28:201–05.
baleen. Anat Rec 298:691–702. Karlsen K. 1962. Development of tooth germs and adjacent
El Adli JJ, Deméré TA. 2015. On the anatomy of the tempo- structures in the whalebone whale (Balaenoptera physalus
romandibular joint and the muscles that act upon it: ob- (L.)). Hvalrådets Skrifter 45:5–56.
servations on the gray whale, Eschrichtius robustus. Anat Ketten DR. 2004. Marine mammal auditory systems: a sum-
Nachtigall PE, Mooney TA, Taylor KA, Yuen MM. 2007. Springer MS, Emerling CA, Fugate N, Patel R, Starrett J,
Hearing and auditory evoked potential methods applied Morin PA, Hayashi C, Gatesy J. 2016a. Inactivation of
to odontocete cetaceans. Aquat Mammal 33:6–13. cone-specific phototransduction genes in rod monochro-
Okazaki Y. 2012. A new mysticete from the upper Oligocene matic cetaceans. Front Ecol Evol 4:61.
Ashiya Group, Kyushu, Japan and its significance to mys- Springer MS, Starrett J, Morin PA, Lanzetti A, Hayashi C,
ticete evolution. Bull Kitakyushu Mus Nat Hist Hum Hist Gatesy J. 2016b. Inactivation of C4orf26 in toothless pla-
A 10:129–52. cental mammals. Mol Phylogenet Evol 95:34–45.
Park JY, An YR, Kanda N, An CM, An HS, Kang JH, Kim Szewciw LJ, de Kerckhove DG, Grime GW, Fudge DS. 2010.
EM, An DH, Jung H, Joung M, et al. 2015. Cetacean evo- Calcification provides mechanical reinforcement to whale
lution: insights from the genome sequences of common baleen -keratin. Proc R Soc B 277:2597–605.
minke whales. BMC Genom 16:13. doi:10.1186/s12864- Thewissen JGM, George J, Rosa C, Kishida T. 2011. Olfaction
015-1213-1. and brain size in the bowhead whale (Balaena mysticetus).
Parks SE, Ketten DR, O’Malley JT, Arruda J. 2007. Mar Mammal Sci 27:282–94.