One of last years greatest surprises was the removal of Hydatellaceae from Poales to Nymphaeales. It might not sound grand, but since Poales is the group of monocots that the grasses, sedges, rushs (and pineapples!) are a part of, and the Nymphaeales are the water lilies, the magnitude of the transfer of this small family of minute, sedge-like aquatics is so more impressing. And even more interesting since the Nymphaeales is one of the prime suspects for the sistership to the rest of the flowering plants! The implication of all this is simply that Hydatellaceae migh be the one of the missing keys needed to solve Darwin’s “abominable mystery“, the grand origin of flowers and flowering plants themselves! When DNA sequences first was used in large scale to resolve evolutionary relationships among plants, the big surprise then was finding the hornwort family, Ceratophyllaceae, at the very base of the Angiosperm tree of life. Soon however, the hornworts lost their isolated position outside all the other Angiosperms, and was replaced by a virtually unknown and thoroughly neglected small shrub from New Caledonia, the now famous Amborella trichopoda. Amborella has since then as best as it can defended its position as sister to the rest of the flowering plants, but what can be ask for from a shrub? From time to time its sistership has been challenged, but no obvious candidate has turned up. One of the challengers have however been Nymphaeales, either alone or together with Amborella, but already small changes in parameters or data sets have restored Amborella as the sole sister. And we are still waiting for the definitive analysis that will show us with high confidence what the basal relationships are. In the meantime, we need to see what the traditional data sets can tell us about the evolution. One possible large step forward was presented in the last issue of Nature, in the paper “Hydatellaceae are water lilies with gymnospermous tendencies” by William Friedman of University of Colorado (Nature 1 May 2008, vol. 453: 94-97). Here Friedman describe the embryology of the female gametophyte (the structure harboring the egg cells and, after fertilization, forms the seed). Not surprisingly, now when we are pretty certain that Hydatellaceae are “water lilies” (albeit seemingly sedgy ones), much of the embryological features correspond to the other water lilies (the female gametophyte of the water lilies and the likewise early-branching Austrobaileyales have a development not found in most other angiosperms, including Amborella), but Hydatellaceae have an additional feature unique among flowering plants: the allocation of nutrients (starch) to the embryo-nourishing tissue prior to the fertilization! All other flowering plants allocate nutrients first after the fertilization, thus avoiding “wasting” valuable nutrients on unfertilized seeds, but the gymnosperms (conifers and allies) don’t. They allocate nutrients before the embryo is formed, and interestingly this is also the case in Hydatellaceae! What the implications of this finding are is not apparent yet, but one obvious possibility is that this is the plesiomorphic (primitive or original) condition among the flowering plants, later changed (lost) in all other flowering plants with the evolution of second fertilization (trigging the onset of the formation of the nutrition tissue), including Amborella, but still present in Hydatellaceae. And the easiest explanation for this pattern is that Nymphaeales is the sister to the rest of the flowering plants, with two “independent” origins of the second fertilization (one on the branch with the other members of the water lilies, one on the branch with all other flowering plants including Amborella). With Amborella as sister to the rest, three “independent” origins are needed, or we need to interpret the pre-fertilization nutrition tissue in Hydatellaceae as a truely unique feature, not really the same as in the gymnosperms (there are differences) but instead evolved from an ancestor with second fertilization, something that is not easily explained. The next logic step from here will be to describe and characterize the genetic system behind this feature, and compare it with the other early branching angiosperms as well as the gymnosperms. The evo-devo people are probably already working hard with it!