INSIGHTS | P E R S P E C T I V E S
PALEONTOLOGY
How land plant life cycles first evolved
Fossils and developmental genetics reveal early changes in the plant life cycle
By Paul Kenrick
T
his year marks the 100th anniversary
of the first in a series of papers on
the biota of a 407-million-year-old hot
spring system that opened a window
onto early life on land (1). The site
near the village of Rhynie in Aber-
deenshire, Scotland, is exceptional because
fossilization occurred in microcrystalline sil-
ica (chert), preserving organisms to the cel-
lular level and shedding light on community
structure and interactions among the plants,
arthropods, fungi, algae, and cyanobacteria.
Recent research on these remarkable fossils
and advances in understanding plant devel-
opmental genetics are beginning to reveal
how major changes in life cycle had an early
influence on the direction of plant evolution.
When the Rhynie Chert fossils were first
reported, the plants, in particular, caused
quite a stir because they seemed to capture
an early stage in the adaptation to life on
land. Most lack key organs familiar in living
species, such as leaves or roots. Moreover,
they are small (<20 cm), with simple bifur-
cating axes that terminate in spore-bearing
sacs. Later research has revealed that their
life cycles also differed in important ways
from those of living species.
Land plants inherited their biochemistry
and cell biology from ancestral green algae,
but their fundamental organs and tissues
evolved on land. Their closest algal rela-
tives have a haplontic life cycle, which typi-
cally features a simple multicellular haploid
sexual phase (a gametophyte) and a unicel-
lular diploid zygote (2). This implies that the
ancestor of land plants was also haplontic,
probably with a filamentous, weakly differ-
entiated multicellular phase (3).
The transition to land entailed changes
in life cycle in which the diploid zygote also
became multicellular (4). This change was
accompanied by substantial somatic devel-
opment, resulting in the evolution of tissues
and organs basic to plants, such as stems,
leaves, roots, a vascular system, stomata
(the minute pores that facilitate gaseous ex-
change), and sex and dispersal organs. The
multicellular diploid phase evolved into a
highly successful spore-producing dispersal
unit (the sporophyte), laying the foundations
Department of Earth Sciences, The Natural History Museum,
London SW7 5BD, UK. Email: p.kenrick@nhm.ac.uk
1538 22 DECEMBER 2017 • VOL 358 ISSUE 6370
of modern plant diversity. Changes in life
cycle thus underpinned the early diversifica-
tion of plants on land, but how such changes
evolved remains a puzzle.
Today, life cycles vary considerably among
major plant lineages (4, 5). In bryophytes
(mosses, liverworts, and hornworts), the spo-
rophyte is a small stalked capsule that is nu-
tritionally dependent on its parent, which is
a cosexual or a female gametophyte of leafy
or crustose type. In vascular plants (lycopods,
ferns, gymnosperms, and flowering plants),
the converse happens. The sporophytes are
large, leafy, free-living plants. In flowering
plants and gymnosperms, the female game-
tophyte is reduced to an ovule typically borne
in a cone or a flower, and the male is reduced
to motile sperm or a minute tube that grows
from a pollen grain. By contrast, in ferns and
lycopods, after a short period of embryonic
development, the two parts of the life cycle
become independent. The sporophyte is large
“...these fossils elucidate
a major early shift in life
cycle that had far-reaching
consequences for plant
evolution.”
and has well-developed leaves, roots, and
vascular system, whereas the gametophyte is
miniscule. It can be surficial and photosyn-
thetic or subterranean and heterotrophic, but
with very limited tissue development.
In phylogenetic terms, the bryophyte life
cycle is thought to be intermediate between
that of the algal ancestors and those of vas-
cular plants (2, 4, 5). Scientists have there-
fore assumed that the life cycles of early
plants would conform to the bryophyte or
the fern and lycopod types, which are con-
sidered ancestral in vascular plants. But the
Rhynie Chert fossils suggest otherwise.
Life-cycle phases are known in varying de-
grees of detail for four of the six fossil plant
species from the Rhynie Chert. Like ferns, the
gametophytes and sporophytes lived as in-
dependent plants, but unlike any living land
plants, tissues such as rooting structures, a
vascular system, and stomata were expressed
in both parts of the life cycle (6). Despite these
similarities, sporophytes and gametophytes
Published by AAAS
were not identical. Although both were axial
and leafless, the gametophytes were smaller.
Overall, habit and size are still poorly un-
derstood for several gametophytes, but one
factor known to influence their size was
the degree of development of their branch-
ing systems (1). Also, in several species, ga-
metophyte axes terminated in an expanded
cup-shaped structure that bore the sexual or-
gans. These differences notwithstanding, the
gametophyte bore much greater similarity to
the sporophyte than it does in living species,
Downloaded from http://science.sciencemag.org/ on December 21, 2017
and it developed tissues of greater diversity.
Neither bryophytes nor ferns, these fossils
elucidate a major early shift in life cycle that
had far-reaching consequences for plant evo-
lution. This shift involved the liberation of
the sporophyte from complete physiological
dependence on its gametophyte (a bryophyte
life cycle) to achieve free-living status (a
vascular plant life cycle). It entailed radical
changes to the sporophyte’s form and func-
tion. The fossils hint at how this transition
might have taken place.
Molecular phylogenetic analysis of liv-
ing species shows that the bryophytes are
basal lineages in the land plant tree of life
(2), implying that the basic features of the
land plant body plan evolved initially in
the gametophyte to create a multicellular
free-living plant (see the figure) (5). Thus,
the shift away from sporophyte nutritional
dependency happened in the ancestors of all
vascular plants. Phylogenetic analyses that
include fossils consistently place the extinct
Rhynie Chert species closer to the vascular
plants than to the bryophytes. Like vascular
plants, they all have free-living sporophytes
with bifurcating axes. Some Rhynie Chert
plants also possessed tracheids, which are
distinctive water-conducting cells found only
in vascular plants. These extinct life-cycle
variants therefore most likely represent an
ancestral condition of the vascular plants (6).
Unlike their living relatives, though, simi-
lar histology in the two parts of the fossil life
cycles implies similar developmental pro-
cesses and functions. This suggests that the
mechanism behind the leap to sporophyte in-
dependence involved co-option by the sporo-
phyte of those tissue systems that enabled the
gametophyte to live as a fully independent
autotroph (6). In this model, the sporophyte
did not evolve novel structures to become
free living and autotrophic; rather, the exist-
ing repertoire of gametophyte developmental
sciencemag.org SCIENCE
 How plant life cycles evolved metophyte has experienced
Plants that lived in the 407-million-year-old Rhynie Chert hot spring system had life cycles that were different from living species. progressive loss of tissues,
beginning with the vascular
system and stomata. One ex-
Ancestral green algae Haploid gametophyte Diploid sporophyte or zygote planation of this divergence
in form between the two
phases of the life cycle is that
Zygote becomes a multicellular sporophyte that is nutritionally dependent on gametophyte the gametophyte evolved
through a persistent, subter-
ranean, mycotrophic phase,
Sporophyte becomes free living and fully autotrophic as seen in some basal groups
Transpiration initiated with stomata change of vascular plants today (14).
Direct fossil evidence for
Gametophyte life cycles in early land plants
passes through is still very sparse. The devel-
subterranean phase oping understanding of the
Rhynie Chert fossils provides
Charophycean algae Bryophytes Rhynie Chert fossils Living vascular plants a model for recognizing life
cycles in less well-preserved
but more abundant com-

Downloaded from http://science.sciencemag.org/ on December 21, 2017

Zygote Sporophyte
processes was redeployed in the sporophyte,
with some limited redeployment in the re-
verse direction (for example, stomata).
Evidence from developmental genetics
lends support to this idea. A growing body of
research points to the recruitment of ancient
genes and gene regulatory networks early in
land plant evolution from a preexisting ga-
metophyte generation to the sporophyte (7).
At the tissue and cellular levels, similar gene
networks regulate root hair development in
angiosperm sporophytes and the develop-
ment of rhizoids (tip-growing rooting cells)
in moss gametophytes (8). Likewise, the de-
velopment of water-conducting cells is con-
trolled by similar transcription factors in the
angiosperm sporophyte and the moss game-
tophyte (9). Polar auxin transport, a key reg-
ulator of sporophyte ontogeny, is essential
to organ and tissue patterning in the moss
gametophyte (10, 11). These observations are
consistent with ancient roles for these de-
velopmental regulators in the gametophyte.
This fits with the Rhynie Chert life cycles, in
which rhizoids and the vascular system are
expressed in both life-cycle phases.
Not all findings of developmental genet-
GRAPHIC: K. SUTLIFF/SCIENCE
Primitive life
cycle exemplifed
by fern
this is regulated by the expression of KNOX2
(KNOTTED-like TALE homeobox gene class
II) in the sporophyte and PRC2 (polycomb
repressive complex 2) complex genes in the
gametophyte (12). In the Rhynie Chert life cy-
cles, repression of sporophyte ontogeny in the
gametophyte and vice versa is less complete.
This implies differences in the downstream
regulation of gene networks by KNOX2 and
PRC2 complex genes in the fossils.
If shifting patterns of gene expression
played an important role in transforming
plant life cycles, they might also have brought
together tissue systems and cell types that
had initially evolved separately. Stomata are
thought to have originated in the nutrition-
ally dependent sporophytes of plants with
bryophyte-like life cycles, in which their main
role was to facilitate spore dispersal; they only
later acquired a role in gaseous exchange in
the vascular plants (13). By contrast, the vas-
cular system and rhizoids probably evolved
separately in the gametophyte (5). The shift
in life cycle leading to the vascular plants
brought rhizoids, vascular tissues, and sto-
mata together in one developmental system
for the first time. Thus, the key components
pression fossil floras. Direct
evidence of life cycles in the
earliest land plants before
the Rhynie Chert fossils is
limited to spores dispersed
in sediments. These provide
tantalizing glimpses. Unlike
modern species in which
most spores are dispersed
as single grains, these ear-
liest types were typically
dispersed in packets of two or four, demon-
strating that the packaging of the products
of meiosis was more diverse than in plants
today (15). Broader interrogation of the fossil
record for further evidence of life-cycle diver-
sity may yield more surprises. j
RE FERENCES AND NOTES
1. D. Edwards, P. Kenrick, L. Dolan, Philos. Trans. R. Soc. Lond.
B 10.1098/rstb.2017.0489 (2017).
2. N. J. Wickett et al., Proc. Natl. Acad. Sci. U.S.A. 111, E4859
(2014).
3. C. F. Delwiche, E. D. Cooper, Curr. Biol. 25, R899 (2015).
4. Y.-L. Qiu, A. B. Taylor, H. A. McManus, J. Systemat. Evol. 50,
171 (2012).
5. R. Ligrone, J. G. Duckett, K. S. Renzaglia, Ann. Bot. 109, 851
(2012).
6. P. Kenrick, Philos. Trans. R. Soc. B 10.1098/rstb.2017.0149
(2017).
7. N. D. Pires et al., Proc. Natl. Acad. Sci. U.S.A. 110, 9571
(2013).
8. T. H. Y. Tam, B. Catarino, L. Dolan, Proc. Natl. Acad. Sci.
U.S.A. 112, E3959 (2015).
9. B. Xu et al., Science 343, 1505 (2014).
10. T. A. Bennett et al., Curr. Biol. 24, 2776 (2014).
11. T. Viaene et al., Curr. Biol. 24, 2786 (2014).
12. J. L. Bowman, K. Sakakibara, C. Furumizu, T. Dierschke,
Annu. Rev. Genet. 50, 133 (2016).
13. J. G. Duckett, S. Pressel, K. M. Y. P’ng, K. S. Renzaglia, New
Phytol. 183, 1053 (2009).
14. D. Haig, Bot. Rev. 74, 395 (2008).
15. D. Edwards, J. L. Morris, J. B. Richardson, P. Kenrick, New
Phytol. 202, 50 (2014).

ics sit comfortably with the early fossil life
cycles, however. In the life cycles of living
plants, one genome gives rise to two distinct
ontogenies. The gametophyte developmen-
tal program must be completely repressed
in the sporophyte and vice versa. In moss,
that regulate transpiration were put in place
in the vascular plants, forming a physiologi-
cal platform of primary importance to their
subsequent diversification.
The sporophyte of vascular plants came
to dominate land floras, whereas the ga-
ACKNOWLEDGME NTS
I am indebted to D. Edwards, L. Dolan, and The Royal Society of
London for funding an interdisciplinary discussion meeting on
“Rhynie Chert—Our earliest terrestrial ecosystem revisited,”
held in March 2017.
10.1126/science.aan2923

SCIENCE sciencemag.org 22 DECEMBER 2017 • VOL 358 ISSUE 6370 1539

Published by AAAS
How land plant life cycles first evolved
Paul Kenrick

Science 358 (6370), 1538-1539.

DOI: 10.1126/science.aan2923

Downloaded from http://science.sciencemag.org/ on December 21, 2017

ARTICLE TOOLS http://science.sciencemag.org/content/358/6370/1538

REFERENCES This article cites 13 articles, 4 of which you can access for free
http://science.sciencemag.org/content/358/6370/1538#BIBL

PERMISSIONS http://www.sciencemag.org/help/reprints-and-permissions

Use of this article is subject to the Terms of Service

Science (print ISSN 0036-8075; online ISSN 1095-9203) is published by the American Association for the Advancement of
Science, 1200 New York Avenue NW, Washington, DC 20005. 2017 © The Authors, some rights reserved; exclusive
licensee American Association for the Advancement of Science. No claim to original U.S. Government Works. The title
Science is a registered trademark of AAAS.