Origin of Leaves
Currently, research indicates that leaves have evolved several times in different lineages. It is possible that flattened, photosynthetic organs that we call "leaves" have evolved at least 6 different times in evolutionary history: leafy liverworts, mosses, clubmosses, horsetails, true ferns, and woody plants). The simplest difference between leaves are the microphylls of the lycophytes, (e.g. clubmosses), and the megaphylls of the ferns and seed plants. There is debate whether fern leaves are homologous to the leaves of horsetails. Also, whether fern leaves are homologous to those found in the woody plants.
Earliest known leaves, by taxonomic group:
Non-vascular plants
Both the mosses and the liverworts have independently evolved leafy forms. These are not considered true leaves because they lack vascular tissue, but are distinctly leaf-like and used for photosynthesis. Some researchers call these structures phyllids. Phyllids are found from the Middle Ordovician, ~465-460 Ma
Above: a leafy liverwort, Scapania
Basal Vascular Plants
The Lower Devonian taxa "Adoketophyton, Stachyophyton and Dibracophyton have fertile leaf-like appendages (sporophylls or bracts) to constitute a strobilar structure with different sterile organs, but they are neither microphyllous plants nor megaphyllous plants" (Hao & Xue 2013)
These sterile organs may have been photosynthetic or filled with chlorophyll during a portion of their development
Right: Reconstruction of the strobilus of Adoketophyton showing leaf-like sporophylls
Lycophytes (e.g. clubmosses)
Baragwanathia is a Late Silurian clubmoss and the earliest known lycophyte to possess microphylls or lycophylls.
These leaves are linear, and like other leaves, they possess a single, unbranched vein in the blade
Above: Compression fossil of Baragwanathia
Eophyllophyton
Eophyllophyton bellum, from the Lower Devonian (Pragian) of China, is the oldest known plant with megaphylls, or true leaves
Megaphylls have multiple, branching veins compared to the leaves in lycophytes
The leaves appear to be flattened shoot systems, and may represent a different origin for megaphyllous leaves
This plant is superficially similar to the progymnosperms
Above: the leaves of Eophyllophyton
Sphenophytes
Pseudobornia is a Late Devonian taxon that produces highly-dissected leaves strictly on higher-order branches
The sphenophytes arose from a leafless ancestor, and probably evolved leaves separately from other plant groups
Some have called these leaves equisetophylls, indicating their unique origin for the horsetail / sphenophyte lineage
The Late Paleozoic Equisetales and Sphenophyllales possessed laminate leaves. Leaves are highly-reduced and non-functional in living Equisetum
Right: The leaves of Pseudobornia
Progymnosperms
Archaeopteris, the first true tree (lignophyte), had frond-like leaves
Earlier members of progymnosperms, such as the aneurophytes, did not have laminate leaves
It is assumed that the leaves of the Archaeopteridales were megaphylls, and share developmental ancestry with the fern and seed plants
Some have called these leaves lignophylls, for the leaves in the lignophyte clade, which may have separate origin from true ferns.
Above: reconstruction of the leaves of Archaeopteris
Ferns
Ellesmeris is an early Late Devonian (Frasnian) fern-like euphyllophyte that evolved the earliest known fern-like megaphylls
This plant was a zygopterid fern, with laminate sphenopteroid pinnules
This plant grew in the understory of Archaeopteris forests
Right: Reconstruction of Ellesmeris
Seed plants
Laceya hibernica (Famennian) has the ealiest known leaf from a seed-bearing plant
Hypotheses proposed for the origin of leaves
Enation concept
This hypothesis by F.O. Bower (1908) proposes the steps for the origin of microphyllous (lycopod) leaves from a leafless ancestor. Some early land plants, such as zosterophylls, had small, leaf-like outgrowths that lacked vascular cells; such outgrowths are called enations.
An interesting early lycopod fossil from the Rhynie chert, Asteroxylon, has enations each with a vascular trace that come up to, but not into, the vascular trace. These may represent a transitional stage between zosterophyll enations, and microphylls in derived lycopods.
This concept pertains mainly to clubmoss or lycophyte evolution. Current thinking proposes that clubmosses evolved leaves separately from other plants, such as ferns and seed plants. Botanists call lycophyte leaves "microphylls", due to their small size. A more definitive feature is that they exhibit a single vascular trace (or vein) that does not branch.
The megaphyllous leaf was hypothesized to be created through some of the "telomic" processes, hypothesized by Walter Zimmermann (1952)
Over-topping: Plants evolve from an isotomous branching habit to pseudomonopodial growth, creating lateral axes that are used for photosynthesis and reproduction.
Planation: The three-dimensional branching systems evolve to grow in a two-dimensional or flattened plane. The lateral bracnhes, mentioned above, would growth in an orientation to intercept as much light as possible.
Webbing: Tissues form in-between the axes of these lateral branches to create a laminate leaf
There is also explanation given to the origin of microphyllous leaves, through a different process
Reduction: Multi-branched shoots become shorter and less branched over time, possibily becoming a single vascularized outgrowth like a microphyll
Sterilization concept
Proposed by Kenrick and Crane (1997), which proposes that the first leaves that evolve (i.e. lycophytes) came from sporangia that were sterilized and repurposed for photosynthesis.
Research by Vasco et al. (2016) may provide evidence for this idea of sterilization as the origin of leaves
They show that there is a common leaf developmental mechanism conserved between ferns and seed plants.
The C3HDZ genes express in lycophyte and fern sporangia, and these genes have conserved expression patterns during initiation of lateral primordia, such as leaves or sporangia.
These researchers hypothesize that there is a deep homology of all leaves and that a sporangium-specific developmental program was coopted independently for the development of lycophyte and euphyllophyte leaves.
Additional Resources
An Analysis of the Telome Concept (Boyer 2008)