Matthew Van Dam
Trigonoscuta & weevil research
I am currently revising a genus of weevils Trigonoscuta, which are restricted to sand dunes . All members are flightless. I am using morphology as well as molecular data to delimit species. I have found of particular use the endopallus (seen in the figures to the right). I hope to have the revision completed in the coming year. I have examined this character system in other weevils as well and it gives a good number of characters to delimit species and is far more diagnostic than the median lobe.
I am also currently revising another flightless dune restricted weevil Miloderes using a similar approach.
Endophalli of Trigonoscuta. A. T. dalei Pierce 1975, lateral view B. T. sp. “Panamint Valley”, lateral view C. T. c.f. holtvillei Pierce 1975, lateral view D. T. sp. “La Poza Grande”, dorsal view E. T. sp. “Rice Dunes”, lateral view F. T. sp. “San Bruno”, lateral view G. T. stantoni Sleeper 1975, dorsolateral view H. T. c.f. sanluisi Pierce 1975, lateral view.
Morphologicla research: In an attempt to fully evaluate each character and see if the endopallus is at least as good as external character in phylogenetic reconstruction, consistency index (CI, measure of homoplasy) values for each character were calculated over all of the equally most parsimonious trees. This was done to give a robust estimate of the homoplasy and signal present in the characters that was not reliant upon a randomly chosen “best” tree. CI values for an entire tree are uninformative for character evaluation because changes in character coding (unless done one at a time) cannot be traced back to any one particular character. In the hopes to actually improve the coding of a morphological character, the approach of picking a single most parsimonious tree from which decisions to include or exclude a character are based could be positively misleading, as there are many alternative trees with different topologies. See the CI mean and range whisker plot below derived from data on Trigonoscuta for each character over all the equally most parsimonious trees.
I am also investigating how sand dunes stack up to oceanic islands as a dispersal barrier. As Trigonoscuta is found in both the Channel Islands, coastal and desert sand dunes this allows for a comparison between the different habitat settings. Long story short, the Channel Islands are no more divergent from one another than the desert sand dunes. See the graph below plotting pairwise genetic distance over geographic distance. Reasons for these patterns are due in part to climate change and sea level change. The northern Channel Islands and perhaps the coastal dunes were connected during the last glacial maximum, these groups show little genetic differentiation. To the right is a reconstruction of coast lines (edge of the red color) during the last glacial maximum. In the phylogeny of Trigonoscuta we find a surprising negative correlation between genetic divergence and change in coast lines. Possible indicating the connected nature of the Pacific Coast sand dunes as a result of coast lines changing and Trigonoscuta tracking the movement of the coastal topography.