Maximum parsimony strict consensus topologies from separate nrITS (88 sequences from 76 individuals, 760 sites, 30% pars. informative) and plastid (71 accessions, 2867 sites, <8% informative) datasets. Bolded species names indicate distylous individuals. Parsimony bootstrap scores are above branches and Bayesian posterior probabilities are below. Incongruence occurs in the placement of O. rosea, and in the order of divergence in the Andean lineages.

This is the border of Rio Grande do Sul (the high campos that sit above the ridge, to the left) and Santa Catarina (everything to the right), between Sao Jose dos Ausentes (RS) and Timbe do Sul (SC). This is where we collected O. calva…

Oxalis calva(?) on the edge of Santa Catarina, about 1200 m. This is a mid from a trimorphic population.

Botanists get excited when an herbarium sheet is so expressive. The bulbs of this look much like the collection we made on top of Morro de la Igreja. O. potamophila?

Some Oxalis bulbs we found in South America from top left: O. perdicaria, O. potamophila; center; O. telmatica; lower left: O. bisecta, O. linarantha.

Bayesian 50% majority-rule consensus topology of combined nrITS and plastid datasets with Bayes Multistate reconstruction of ancestral geographic states. Pie graphs represent the distribution of posterior probabilities of ancestral states for each node. Colors represent the geographic coding (Red: Patagonia; Green: SE South America; Purple: Andes; Turquoise: Central America and southern Mexico; Orange: MTVB and north; Dark Blue: Africa). Provinces/departments/states are shaded on the maps based on the distributions of taxa represented by the phylogeny, based on data from Denton (1973) and Lourteig (2000). The widely-distributed weeds O. debilis and O. latifolia are not mapped.

Morphology and geography of the American bulb-bearing Oxalis. Colors represent the geographic coding we use when reconstructing ancestral geographic states (Red: Patagonia; Green: SE South America; Purple: Andes; Turquoise: Central America and southern Mexico; Orange: MTVB and north; Dark Blue: Africa). Provinces/departments/states are shaded based on the number of species represented in each of the sections Articulatae, Palmatifoliae, Pseudobulbosae, and Ionoxalis—those with greater species diversity are more saturated with color. The widely-distributed weeds O. debilis and O. latifolia are not mapped. Representatives of each section are depicted at the same scale. Data from Denton (1973) and Lourteig (2000); illustrations from Rose (1906), Knuth (1930), Denton (1973), and Lourteig (2000).

Well-pressed flowers of oxalis potamophila. A short.

A tristylous mid; from Calli, Marian, Anna, and Josh.

Distribution of American bulb-bearing Oxalis collections (white dots) and precipitation seasonality (aseasonal-green, extremely seasonal precipitation-red).  Oxalis collections are from georeferenced herbarium specimens, precipitation seasonality is from the WORLDCLIM dataset.

I’m thinking about rescaling the precipitation seasonality dataset using latitude, so that areas where precipitation falls as snow for several months are re-scored as being more seasonal.  This would raise the values for areas like Tierra del Fuego and northern North America, perhaps better reflecting the accessibility of precipitation.

Submitted a National Science Foundation Doctoral Dissertation Improvement Grant today. This is the most interesting figure…

The American bulb-bearing Oxalis exhibit dramatic variation in underground storage morphology along two fairly easily measurable axes: X) number of bulbscales per mm of stem axis (how condensed the bulb is), and Y) number of bulblets produced per growing season (how much allocation to asexual propagation).

Distribution of Oxalis section Pseudobulbosae by state/province/dept.  The group is most diverse in southern Brazil.  Data are from Lourteig, 2000, and surprisingly, have no collections mentioned for Uruguay.  The members of this section may belong in Ionoxalis…

Estelita-Teixeira, 1982

Oxalis bulb morphology: CT Scans