Document Type


Publication Date

Fall 11-15-2004


Arceuthobium divaricatum is a dwarf mistletoe that infects the piñon pines of the southwestern United States and Baja California, Mexico. According to packrat midden data, A. divaricatum and its primary host, P. edulis, were restricted to two isolated glacial refugia during the last glacial maximum (LGM). I examined chloroplast DNA (cpDNA) sequence variation in both A. divaricatum and P. edulis populations across their ranges to test the hypothesis that genetic differentiation occurred between the two glacial refugial populations for both species and that the piñon hosts had an impact on the population genetic structure of A. divaricatum. For P. edulis, I used portions of the rbcL and MatK regions in the chloroplast to construct a phylogeny of P. edulis populations and of other piñons including P. discolor, P. cembroides, P. quadrifolia, P. johannis, P. remota, P. monophylla, P. monophylla X P. edulis hybrids, P. californiarum and var. fallax, and P. californiarum var. fallax X P. .frav P. edulis hybrids. Pinus edulis populations were divided into two clades, consistant with the hypothesis of two glacial refugia. Furthermore, P californiarum var. fallax was not differentiated from P. edulis supporting previous literature that suggests these two are not distinct species. I used the trnS-trnfM, trnR-G, and trnG-S intergenic regions and the trnG gene to construct a phylogeny of A. divaricatum populations and to examine distribution of the genetic variation. An analysis of molecular variance (AMOVA) revealed a significant partitioning of the genetic variation (p<0.001) among geographic regions. The phylogeny revealed three distinct clades and one grade. These data support the hypothesized locations of refugia deduced from packrat midden data and suggest that there were at least two other refugia, one in the Baja California / California region and one in the Guadalupe Mountains. Employing a molecular clock, the divergence dates are consistent with isolation during the LGM except for the population in the Guadalupe Mountains which diverged at an earlier date. Finally, I combined all data to examine whether the genetic variation in A, divaricatum was influenced by host type and specifically, whether host race formation had occurred. An AMOVA revealed that 12% of the variation was partitioned among host groups, however this was not significant (p=0.06). A phylogeny revealed that only A. divaricatum populations on P. californiarum var. fallax X P. edulis hybrids were monophyletic suggesting the possibility of host race formation in these populations.