Giant reed (Arundo donax) is an invasive weed that is native to the Old World. Tens of thousands of hectares of riparian habitat in the Rio Grande Basin (RGB) in Texas and Mexico have been heavily affected by invasions of Arundo. Additionally, many other watersheds across the southwestern United States have also been affected. Giant reed is being targeted for biological control because it displaces native vegetation and consumes water that could potentially be used for agricultural and municipal purposes, especially in areas with limited access to water. Finding the best-adapted insects for biological control involves locating the origin(s) of this plant. To narrow down the proximal source(s) of invasion of giant reed in the RGB, 10 microsatellite markers were developed. An analysis of 203 Old World and 159 North American plants, with an emphasis on the RGB, indicated a reduction in the allelic diversity in the introduced range compared with the Old World. Clonal assignment, neighbor joining, principal coordinates analyses, and STRUCTURE analyses were consistent and implied multiple introductions in North America, with one (likely clonal) lineage responsible for the invasion of the RGB, northern Mexico, and other parts of the southwestern United States. Although no identical matches with the RGB lineage were found in the Old World, several close matches were found on the Mediterranean coast of Spain.

The Old World grass, Arundo donax L. (giant reed), is a serious invader of California riparian areas, and its purported ecosystem impacts led to its consideration as a target for biological control development. However, the herbivore complex in the Arundo adventive range has not been characterized, so there is little information regarding insects that may hinder biological control efforts by interfering with the release of new agents or that could be promoted as augmentative biological control agents if they have a substantial impact on the target weed. Here we report the results of surveys in California, with emphasis on three presumably non-indigenous insects that inflict significant damage to the host. One is a shoot-boring wasp, Tetramesa romana (Walker) (Hymenoptera: Eurytomidae), with a range limited to southern California and that damages shoots generally less than 1 cm in diameter. A shoot fly, Cryptonevra sp., is also associated with shoot damage and often mortality of secondary stems. A third herbivore, the aphid Melanaphis donacis (Passerini), is widespread in the southern and central parts of the state but has less apparent impact to the host. T. romana and Cryptonevra sp. are currently candidates for biological control development and introduction from overseas locations. Their established presence in California suggests that efforts could be revised to focus on documentation of host ranges and impacts under field rather than in quarantine conditions, in anticipation of future re distribution in North America.