= Solution
\b[The innovation is the integration of male and female reproductive organs on one developmentally coordinated, usually determinate shoot. Its origin is best understood as a reorganization of older reproductive programs, although the exact extinct precursor remains uncertain.] A <bisexual flower> bears both \x[stamens]{magic} and <carpels>; this differs from <monoecy>, in which separate male and female structures occur on one individual. Many living <angiosperms> have unisexual <flowers>, so bisexuality is an inferred ancestral feature and a recurrently modified condition, not a definition satisfied by every living member.
The starting materials already existed in ancestral <seed plants>: <microsporangia> producing the male lineage, <ovules> retaining the female lineage, protective structures and reproductive <meristems>. Living <gymnosperms> provide comparisons of these functions, but no extant cone can simply be declared the direct ancestor of the <flower>. A bisexual organization must establish male and female organ identities in distinct positions, integrate their maturation and regulate the growth of a common reproductive axis. Enclosure of <ovules> within <carpels>, a receptive surface for <pollen>, and the elaboration of the <perianth> are related innovations, but their historical order cannot be inferred merely by listing their present-day functions.
Developmental genetics shows how a common axis can generate different reproductive organs. In the <ABC model of flower development>, overlapping regulatory activities specify different organ types. For the reproductive regions, B- and C-class activity together, with suitable cofactors, specifies <stamen> identity, while C-class activity without B-class activity specifies <carpel> identity. In <Arabidopsis thaliana>, the relevant programs involve <APETALA3>, <PISTILLATA> and <AGAMOUS>, with <SEPALLATA> activities required for normal specification of the inner organs. Higher-order loss of <SEPALLATA> function transformed their identities, providing experimental evidence that the spatial code depends on additional regulatory partners.<B and C floral organ identity functions require SEPALLATA MADS-box genes by Soraya Pelaz and others>[Nature 405, pp. 200–203; sepallata multiple mutants and floral organ identity]{ref} The simple A/B/C lettering summarizes functional experiments in model <plants>; its exact domains and functions are not universal across all <flowers>.
Many of these regulators are <MADS-box transcription factors> acting in interacting complexes. <gene duplication>, divergence of <protein> interactions, changes in target <genes> and <cis-regulatory evolution> can therefore generate new combinations and spatial boundaries without creating male and female development from nothing. Comparative work found related B-class activity in <gymnosperms>, including experimental effects of a <Gnetum> regulator expressed in <Arabidopsis thaliana>.<On the origin of class B floral homeotic genes: functional substitution and dominant inhibition in Arabidopsis by expression of an orthologue from the gymnosperm Gnetum by Kai-Uwe Winter and others>[The Plant Journal 31, pp. 457–475; heterologous expression and effects on floral B function]{ref} This supports continuity in reproductive regulatory machinery. Heterologous activity does not show that a gymnosperm <gene> had the same entire function in its native ancestor, or that the modern angiosperm network was already assembled.
One plausible developmental route is spatial subdivision of a previously unisexual axis. In an <out-of-male hypothesis>, restriction of B-class activity within an originally male reproductive structure allows a distal region with C-class but no B-class activity to express female identity, while the proximal region retains male identity. Conversely, recruitment of male activity into part of an initially female structure is developmentally conceivable. These are comparative models of changing organ identities; contemporary homeotic transformations demonstrate accessibility of a developmental outcome, not the sequence of events that actually occurred in an extinct lineage.
The <mostly male theory> makes a more specific historical proposal: an ancestral predominantly male structure acquired ectopic ovules and ultimately a bisexual organization. Its arguments combine morphology, fossils and hypotheses about reproductive regulators, including the history of <LEAFY> and <NEEDLY> homologues.<The mostly male theory of flower evolutionary origins: from genes to fossils by Michael W. Frohlich and David S. Parker>[The Botanical Review 66, pp. 143–171; proposed origin of bisexuality and links between gene histories and reproductive morphology]{ref} It should not be equated with every model involving a shift in B-class expression. Nor does the loss or retention of one duplicated regulator uniquely determine the ancestral cone morphology: expression domains can evolve, <gene> histories can be difficult to resolve, and the relevant intermediates may be extinct. Another possibility is condensation and reorganization of an axis bearing separate reproductive units. Fossil anatomy is needed to discriminate organ-level and whole-axis homologies.
<ancestral state reconstruction> across living <angiosperms> supports a bisexual ancestral <flower>, while leaving uncertainty about several details of its organization.<The ancestral flower of angiosperms and its early diversification by Hervé Sauquet and others>[Nature Communications 8, article 16047; reconstruction of the crown-angiosperm flower and uncertainty across floral characters]{ref} This concerns the last common ancestor of living <angiosperms>, not necessarily the very first transitional reproductive shoot on their stem lineage. Different characters need not have arisen simultaneously. Fossils can constrain plausible combinations and their ordering, whereas developmental experiments reveal which reorganizations can work. Neither line of evidence alone establishes a complete historical sequence.
The adaptive opportunities of integration are substantial but conditional. A shared <perianth> or floral reward can attract a visitor that both removes <pollen> and deposits it on receptive <stigma (botany)> surfaces. Spatially coordinated organs can make transfer more precise and reduce the construction cost of maintaining separate displays. Bisexuality also permits flexible allocation between male and female reproduction as resources or mating opportunities change. When cross-pollination is unreliable, some self-compatible <flowers> can obtain <reproductive assurance> through self-fertilization. That benefit depends on compatibility and timing, however: bearing both organ types does not itself guarantee selfing.
Integration also creates conflicts. Transfer of a <plant>'s own <pollen> to its receptive surfaces can reduce outcross <pollen> export, block opportunities for outcrossing, or produce <inbreeding depression>. <herkogamy> separates organs spatially, <dichogamy> separates their maturation in time, and <self-incompatibility> can reject genetically unsuitable <pollen>. These mechanisms show how the benefits of a shared shoot can be retained while limiting self-fertilization; they need not all have accompanied its initial origin. Later transitions to separate sexes can be favored in other ecological or genetic settings.
Tests of adaptive significance should compare <pollen> export, receipt, <seed> set and offspring performance, rather than count visits alone. Manipulate organ position or timing while controlling display size and reward; compare cross-pollination and autonomous selfing treatments, with emasculation and hand-pollination controls where appropriate. A diversification analysis must also account for shared ancestry and other innovations: the coincidence of bisexual <flowers> and the success of <angiosperms> does not isolate bisexuality as the cause of their <species richness>. \b[A defensible account combines inherited reproductive regulatory modules, new spatial coordination and conditional reproductive advantages, while treating the precise precursor and causal role in diversification as hypotheses.]
Back to article page
