[7] The roots broaden to a swollen base up to 5 mm (0.2 in) wide where they attach in clusters to a bulb-like, underground rhizome characteristic of most quillwort species, though a few (e.g.

This swollen base also contains male and female sporangia, protected by a thin, transparent covering (velum), which is used diagnostically to help identify quillwort species.

[8] They also possess a vestigial form of secondary growth in the basal portions of its cormlike stem, an indication that they evolved from larger ancestors.

Some aquatic species do not have stomata and the leaves have a thick cuticle which prevents CO2 uptake, a task that is performed by their hollow roots instead, which absorb CO2 from the sediment.

[7] CAM is normally considered an adaptation to life in arid environments to prevent water loss with the plants opening their stomata at night rather than in the heat of the day.

The development of vascular tissue and subsequent diversification of land plants coincides with the increased dominance of the sporophyte and reduction of the gametophyte.

Isoetes, as members of the Lycopodiopsida class, are part of the oldest extant lineage that reflects this shift to a sporophyte dominant lifecycle.

[18] These spores are highly ornate and are the primary way by which species are identified, although no one functional purpose of the intricate surface patterns is agreed upon.

Isoetes sporophytes solve this problem by provisioning starches and other nutrients to the spores as an energy reserve for the eventual gametophytes.

[24] Another critical element of dispersal is the observation that in some species of Isoetes, the outer coat of megaspores have pockets that trap microspores, a condition known as synaptospory.

[26] Isoetes may mitigate this issue via microspores stuck to megaspores, greatly increasing the possibility of successful fertilization upon dispersal.

The first critical monograph on their taxonomy, written by Norma Etta Pfeiffer, was published in 1922 and remained a standard reference into the twenty-first century.

Vegetative characteristics commonly used to distinguish other genera, such as leaf length, rigidity, color, or shape are variable and depend on the habitat.

[30] The earliest fossil that has been assigned to the genus is †Isoetes beestonii from the latest Permian[31] of New South Wales, Australia, around 252 million years ago.

[2] Isoetites rolandii from the Late Jurassic of North America has been described as the "earliest clear example of a isoetalean lycopsid containing all the major features uniting modern Isoetes", including the loss of the elongated stem and vegetative leaves.

Wood et al (2020) asserted there to be no morphological features that define the major clades within Isoetes, and no fossils are known that can be definitively assigned to the crown group.

[2] While Wood et al. suggested a young origin dating to the early Cenozoic based on molecular clock estimates[2], the results were questioned by Wikström et al. (2023) who regarded the molecular clock as providing no firm evidence for the origin time of the genus, which could date to the Mesozoic or even the late Paleozoic, depending on the calibration method used.