Ploidy race distributions since the Last Glacial Maximum in the North American desert shrub, Larrea tridentata

• 1 A classic biogeographic pattern is the alignment of diploid, tetraploid and hexaploid races of creosote bush (Larrea tridentata) across the Chihuahuan, Sonoran and Mohave Deserts of western North America. We used statistically robust differences in guard cell size of modern plants and fossil leaves from packrat middens to map current and past distributions of these ploidy races since the Last Glacial Maximum (LGM).
• 2 Glacial/early Holocene (26–10 ¹⁴C kyr bp or thousands of radiocarbon years before present) populations included diploids along the lower Rio Grande of west Texas, 650 km removed from sympatric diploids and tetraploids in the lower Colorado River Basin of south‐eastern California/south‐western Arizona. Diploids migrated slowly from lower Rio Grande refugia with expansion into the northern Chihuahuan Desert sites forestalled until after ~4.0 ¹⁴C kyr bp . Tetraploids expanded from the lower Colorado River Basin into the northern limits of the Sonoran Desert in central Arizona by 6.4 ¹⁴C kyr bp . Hexaploids appeared by 8.5 ¹⁴C kyr bp in the lower Colorado River Basin, reaching their northernmost limits (~37°N) in the Mohave Desert between 5.6 and 3.9 ¹⁴C kyr bp .
• 3 Modern diploid isolates may have resulted from both vicariant and dispersal events. In central Baja California and the lower Colorado River Basin, modern diploids probably originated from relict populations near glacial refugia. Founder events in the middle and late Holocene established diploid outposts on isolated limestone outcrops in areas of central and southern Arizona dominated by tetraploid populations.
• 4 Geographic alignment of the three ploidy races along the modern gradient of increasingly drier and hotter summers is clearly a postglacial phenomenon, but evolution of both higher ploidy races must have happened before the Holocene. The exact timing and mechanism of polyploidy evolution in creosote bush remains a matter of conjecture.

     

Reticulate evolution in Thuja inferred from multiple gene sequences: implications for the study of biogeographical disjunction between eastern Asia and North America

The eastern Asia-North America disjunction is one of the most interesting biogeographical patterns, but its formation is still in much debate. Here nucleotide sequences of five cpDNA regions, nrDNA ITS and two low-copy nuclear genes (LEAFY, 4CL) were employed to reconstruct the phylogeny and to explore the historical biogeography of Thuja, a typical eastern Asia-North America disjunct genus. High topological discordance was observed between chloroplast and nuclear gene trees, even between different nuclear gene trees, suggesting that Thuja could have a reticulate evolutionary history due to multiple interspecific hybridization events. The eastern Asian species Thuja koraiensis might have obtained its chloroplast genome from the eastern North American species T. occidentalis by chloroplast capture, while the western North American species T. plicata is very likely to have inherited a recombinant cpDNA. Based on the phylogenetic analysis of multiple genes, DIVA-reconstruction of the distribution history, molecular clock estimation and fossil data, we inferred that Thuja could have originated from the high-latitude areas of North America in the Paleocene or earlier with subsequent expansion into eastern Asia through the Bering Land Bridge. The two eastern Asia species T. standishii and T. sutchuenensis have a sister relationship, and their split could have occurred in the Oligocene or early Miocene. In the present study, the selection of molecular markers in biogeographic studies was also discussed. Since most previous studies on the eastern Asia and North America disjunction are based on uniparentally inherited cpDNA and (or) directly sequenced nrDNA ITS data, the historical reticulate evolution in the studied groups might have been underestimated. Therefore, we suggest that multiple genes from different genomes, especially low-copy nuclear genes, be used in this research area in the future.