Palynology from A to Z: A as in Aponogeton

One thing, full-blood scientists usually forget, is to advertise their work. Usually because they lack the time. I have plenty and started a series of daily threads on Twitter advertising palynological research. But my reach there is miniscule and the half-life of tweets is extremely short. Hence, this post series.

A as in Aponogeton

Grímsson F, Zetter R, Halbritter H, Grimm GW. 2014. Aponogeton pollen from the Cretaceous and Paleogene of North America and West Greenland: Implications for the origin and palaeobiogeography of the genus. Review of Palaeobotany and Palynology 200:161–187. [Twitter-thread] 

My colleagues in Vienna found something that appeared to be Aponogeton pollen grains in late Cretaceous/Palaeogene strata of North America and Greenland, far outside the modern distribution area.

Oldest fossil record (and one of the few) of Aponogeton (Grímsson et al. 2014, plates I and II). Left panels: Aponogeton harryi sp.nov. from the lower Campanian Elk Basin, Wyoming (82–81 Ma); right: A. hareoensis sp.nov. from the Lutetitian of Qeqertarsuatsiaq, Disko Bay, western Greenland (44–40 Ma).

Aponogeton (Alismatales:Aponogetaceae) is a tropical-subtropical genus of wide-spread monocot water plants that has so far attracted relative little systematic attention because it is a can of worms. A first molecular phylogeny was put forward by Les et al. (2005), who also included a morphological matrix, focussing on the Australian species. They found quite puzzling molecular differentiation patterns indicating that basically every morphological trait of the genus is extremely homoplastic (i.e. variable and poorly sorted within the molecular-defined lineages). How poorly sorted morphological traits are, can be seen in this visualisation of Les et al.'s morphomatrix mapping the pollen types described in Grímsson et al. (2014).

Grímsson et al. (2014), fig. 2. A – neighbour-net based on Les et al.'s morphological matrix. The boxy structure illutrates the poor sorting of morphological traits. Roman numbers indicate main molecular clades B – Les et al.'s molecular tree reduced to inter-clade relationships.

Lacking any high-quality imaging of Aponogeton pollen, my Vienna friends had to first reinvestigate the pollen of the modern species to assess how diagnostic pollen morphs are for the intergeneric lineages. In course of this, we also expanded Les et al.'s morphomatrix to all species of Aponogeton.

Grímsson et al. (2014), fig. 3 – neighbour-net based on our extended morphological matrix including all described species, many of which were palynologically studied for our paper.

Aponogeton species hybridise as soon as they come into contact. Being a water plant, they are dispersed by migratory birds, with isn't helpful either. Les et al. relied exclusively on simple trees and combined nuclear and plastid data to increase resolution although some of their samples show strongly conflicting signals. Hence, I re-analysed their data to map our pollen types. And we found quite a nice fit.

A neighbour-net based on genetic distances (nuclear-encoded ITS; Grímsson et al. 2014, fig. 4; fig. 5 showing the same for Les et al.'s plastid data)

With respect to the pollen, reaching deep into the past of the Aponogeton lineage, we came up with this synthesis of pollen evolution through space and time.

Abbrev.: AFR = continental Africa; GL = Greenland; MG = Madagascar; SA = South Africa; SAS = South Asia; SEA = Southeast Asia; SWA = temperate southwestern Australia; TAA = tropical Australasia; WNA = western North America (extinct)

The paper further includes palaeogeographic maps showing the fossil record of Aponogeton and putative dispersal routes. Each pollen type of extant species is figured, too, in total the paper includes eight highest quality pollen plates.

Epilogue
Sometime later, a Chinese research group became interested in the genus, contacted a colleague in Munich for assistance and we ended up publishing a molecular phylogenetic paper including most species of the genus (Chen et al. 2015). Adding the much-needed African species, Les et al. had no access to. We tested different dating and biogeographic scenarios also with respect of the 2014 pollen paper.

Chen et al. (2015), fig. 2b/c showing biogeographic ML ancestral state reconstructions on dated trees including fossil pollen as additional, age-controlled tip taxa (these days, there are fancier methods to do this, however, the results are pretty much the same)

Unfortunately, by adding more species, we added more potential past reticulations. Pollen type 2 turned out to be not lineage-specific but scattered across the entire genus, and pollen type 5 was shared by two, genetically distinct clades, and we had some highly intriguing species described from Madagascar by an ingenious taxonomist in Munich, Josef Bogner, where nothing matched up.

Chen et al. (2015), fig. 2 based on combined nuclear and plastid data

Not inexplicable, pollen show usually the most conservative traits of a plant, while genetics react were strongly to hybridisation. Finding a "wrong" pollen in a species of Aponogeton may indicate this species is the product of ancient lineage crossing but its genetics don't show this signal anymore because of backcrossing. In fact, we didn't observe any strong incongruence between the nuclear and plastid signals, but all our material was from material collected in the wild (the herbarium in Munich has a unique collection thanks to Dr Bogner), while Les et al. included cultivated material. But to make a proper call, one would have needed to study the pollen from exactly the samples that one sequences. And nobody was paid to do such highly interesting research.

Personally, I think that in all cases where the pollen didn't match up with the genetics, as they mostly do (see above), it's because of ancient hybridisation, and that pollen type 1 and 2 are the ancestral pollen types of (modern) Aponogeton that survived in some species later radiations. Genetically, the modern-day species don't appear to be very old, but the lineage leading to them, is. I wouldn't be surprised if one finds more Aponogeton pollen in the late Cretaceous and Palaeogene that may again look a bit different or match these two but are from extinct sister lineages: Aponogeton is the only genus of the Aponogetaceae (well, actually its two, A. hexapetalus would deserve a genus on its own).

Further references

• Chen L-Y, Grimm GW, Wang Q-F, Renner SS. 2015. A new phylogeny for the aquatic family Aponogetonaceae, combined with northern hemisphere fossils, rejects the hypothesized Australian origin of the family. Molecular Phylogenetics and Evolution 82:111–117. 
• Les DH, Moody ML, Jacobs SWL. 2005. Phylogeny and systematics of Aponogeton (Aponogetonaceae): The Australian species. Systematic Botany 30:503–519.