(Plant annual), growth sympodial, rhizome short, erect; chemistry?; mycorrhizae 0; cuticular waxes 0; leaves linear, with a single vein, margins entire; plant monoecious; perennial taxa sympodially branched/inflorescence axillary, scapose (sessile), capitate, with involucral bracts; flowers very small, monosymmetric by reduction; P 0; staminate flowers: A 1, filament long, slender, articulated with the anther; endothecium 0; tapetal cells?; pollen with spinules; carpelate flowers: pedicel articulated; G 1, three vascular bundles equidistant (2, 1), style 0, stigma penicillate [hairs as rows of plump cells]; extragynoecial compitum 0; ovule pendulous, anatropous, apotropous, micropyle bistomal, parietal tissue ?0, nucellar cap +/0, suprachalazal nucellar tissue massive; (two embryo sacs developing); fruit smooth to papillate, follicular and splitting into three valves, or ?achenial; testa smooth or reticulate, exotestal cells in files, (with sinuous anticlinal walls), other layers collapsed, tanniniferous; embryo barely differentiated, suspensor unicellular; n = 7, x = 7 (?8), (chromosomes holocentric), nuclear genome [1 C] (0.064-) 1.686(-44.679) pg/1335-1350 Mbp; seedling - see below.

Age. It has been estimated that crown-group Hydatellaceae are (23.4-) 19.1, 17.6(-14.7) Ma (Iles et al. 2014).

Hydatellaceae may be represented in the pollen record in rocks from the Isle of Wight that are some 130 Ma (Hoffmann & Zetter 2010), while the pollen Monosulcites riparius in ca 75-70 Ma rocks from Eastern Siberia has been identified as that of Trithuria - and this of course disagrees with the age in Iles et al. (2004).

Hydatellaceae are rather small, more or less caespitose and often annual aquatic (submerged or not) herbs with linear leaves that have only a single vein, and capitate and usually scapose inflorescences bearing minute, very reduced flowers; the penicillate stigmas with beaded hairs are distinctive.

Evolution: Divergence & Distribution. For the evolution and biogeography of the family - crown Hydatellaceae are certainly not Gondwanan in age - see Iles et al. (2014). Trithuria konkanensis (India) and T. lanterna (N. Australia) diverged (1.3-) 0.76(-0.24) Ma and evolution in the genus seems in general to be pretty active (e.g. Marques et al. 2016 and references).

The inflorescence is described as being cymose and capitate, although bractless and with highly reduced flowers, i.e., it is a sort of pseudanthium, although alternative interpretations are possible (Rudall et al. 2007a, 2009a; Rudall 2010; Rudall & Bateman 2010; Baczynski & Cla?en-Bockhoff 2023); for the sympodial growth of perennial species, see Sokoloff et al. (2009a). The pedicels seem at least sometimes to be articulated. Early work suggested that the carpels might be initiated outside the stamens, and this has been confirmed (Rudall et al. 2007a); staminate flowers are the first to be initiated in the cymose inflorescence, and the result is that several "carpels", developing centrigugally, surround a few "stamens". How the reproductive structures are to be interpreted, whether flowers or inflorescences, remains unclear, and Sauquet et al. (2017) reasonably elected not to include the family in their reconstruction of the morphology of the ancestral angiosperm flower. Trithuria sects Hydatella and Altofinia are considered by Romanov et al. (2023) to have berries of the Schisandra type, in which they also include the fruits of Cabombaceae.

Pollination Biology & Seed Dispersal. For reproductive ecology - wind-, self-, or hyphohydrous pollination - see M. L. Taylor et al. (2010); the pollen tubes grow down the multicellular stigmatic hairs under the cuticle (Prychid et al. 2011).

Genes & Genomes. For chromosome numbers, see Marques et al. (2016). The genome of Trithuria submersa shows signs of polyploidy (n = 28), and meiosis during microsporogenesis has a number of odd features, and it is possible that (some of) the chromosomes, perhaps belonging to one of the genomes, are holocentric (Kynast et el. 2014).

Chemistry, Morphology, etc.. The sieve tube plastids were reported as having triangular proteinaceous inclusions, but the inclusions appear to be of the starchy type as are more to be expected in this part of the tree (Tratt et al. 2009). There is some variation in the epidermis of the root and whether or not root hairs are produced (Sokoloff et al. 2008a). Hairs with possible apical secretory cells are known only from the inflorescences.

The inflorescence is described as being cymose and capitate, although bractless and with highly reduced flowers, i.e., it is a sort of pseudanthium, although alternative interpretations are possible (Rudall et al. 2007a, 2009a; Rudall 2010), indeed, how the reproductive structures are to be interpreted, whether flowers or inflorescences, is unclear. Early work suggested that the carpels might be initiated outside the stamens, and this has been confirmed (Rudall et al. 2007a); staminate flowers are the first to be initiated in the cymose inflorescence (see also Begoniaceae). Perennial Hydatellaceae have a cymose branching pattern (Sokoloff et al. 2009).

The fruit often opens along three lines as the three vascular bundles separate from the rest of the pericarp (see also Sokoloff et al. 2013a). Both integuments have two cell layers; the operculum is formed from enlarged cells of the inner integument. Starch deposition in tissues that will become perisperm begins before fertilization (Friedman 2008a).

There is some disagreement over the interpretation of the embryo. Tuckett et al. (2010: discussion of "ancestral" embryo type for angiosperms must include Amborellaceae, at least; see also Sokoloff et al. 2014) found that the embryo differentiated and the shoot and root appeared only after germination began. Tillich et al. (2007) compared seedling morphology with that of a monocot, describing collar rhizoids, a coleoptile, two cotyledonary sheath lobes, and a haustorium. Sokoloff et al. (2008a) suggested that the sheathing structure with its bilobed apex that is found in some species could be interpreted as two more or less completely connate cotyledons; the rest of the embryo is attached to the sheathing structure. In some taxa there is apparently no sheathing structure at all, only a haustorial lateral outgrowth (cotyledonary) that goes into the rest of the seed, the rest of the cotyledon being photosynthetic, so the seedlings are simultaneously both phanero- and cryptocotylar - c.f. some monocots (Sokoloff et al. 2013b), and Sokoloff et al. (2008a, 2014) suggested that Hydatellaceae showed how monocot-like embryos/seedlings might have originated. Some species are phanerocotylar (Sokoloff et al. 2013b). Both Tillich et al. (2007) and Sokoloff et al. (2008a) examined largely the exterior morphology of the embryo, neither looked in any detail at anatomy (c.f. Friedman et al. 2012: superb micrographs; Sokoloff et al. 2014); for a discussion on germination, see Baskin and Baskin (2018).

Additional information is taken from Hamann (1998: general), Sokoloff et al. (2008b: family monograph, 2011: important literature review, 2019b: species limits), Cutler (1969: vegetative anatomy), Remizowa et al. (2008b: pollen), Hamann (1975), Rudall (1997) and Rudall et al. (2008), both embryology and Hamann et al. (1979), Cook (1983) and Rudall et al. (2009b), all seeds and germination.

Previous Relationships. Hydatellaceae had long been thought to be monocots, largely because of their superficial similarity to Centrolepidaceae (= Restionaceae). Both groups are very reduced morphologically, and indeed Hydatellaceae have been misidentified as Centrolepidaceae. It was unclear if the gynoecium of Hydatellaceae was 1- or 3-carpelate, and since the fruits of Trithuria (= Hydatella) opened by three valves, they looked rather monocot-like. The combination of characters in Hydatellaceae was recognised as being unique to that group, and it made them very distinctive within monocots as a whole (e.g. Hamann et al. 1979; Dahlgren et al. 1985).