Hybridization network for the system of woody plant gene pools in the United States

Abstract

Hybridization generates similarities among gene pools. This structure can be visualized and analyzed at the systems level using networks. Here we construct a network of the 315 woody plant species native or naturalized in the United States using data compiled by the Forest Service of the US Department of Agriculture (USFS). Each species is represented by a node in the network whose size is proportional to a recent census for live stems in the continental United States. Each of the 416 links between node pairs represents evidence for hybridization compiled from the USFS manual Silvics of North America. The total network resolved into 100 separate connected components or clusters (mean size, 3.15 species), with 44% of species linked to at least one other. Betula had the largest component (18 species) following by the separate Quercus clusters (17 red oaks and 16 white oaks); Q. velutina was the most genetically connected woody plant in the continental US. The number of species held together per component (i.e. size) scaled as a power–law albeit a slightly truncated one. The truncation suggests there are fewer than expected hybridizing species within the large woody genera of plants in the US.

Keywords:

• complexity
• graph theory
• hybrid
• power–law scaling
• systems biology

Acknowledgements

We thank Benedictine University for computer support of this project, the US Forest Service for access to their data, and two anonymous reviewers for their comments.