Relative performance and impacts of the psyllid Arytinnis hakani (Hemiptera: Psyllidae) on nontarget plants and the target weed Genista monspessulana (Fabales: Fabaceae).
No-choice tests can help select weed biological control agents with a high degree of host specificity but may exclude potentially effective agents that can develop on nontarget plants under laboratory conditions. The actual amount of damage to nontarget plants often goes unstudied, even though agents may exploit nontarget plants without inflicting significant harm. Furthermore, tests typically assess whether prospective agents can complete one generation on nontarget plants, and rarely examine whether agents are likely to persist on the nontarget plants over the long term. Pre-release assessments that occur over multiple generations of the agent could help determine whether prospective agents pose a threat to nontarget plants under field conditions. This study focused on the psyllid Arytinnis hakani Loginova (Hemiptera: Psyllidae), which is under consideration in California for release against the invasive shrub French broom, Genista monspessulana (L.) LAS Johnson. We examined the host suitability of seven nontarget Lupinus spp. (Fabales: Fabaceae) for the psyllid using no-choice tests, and assessed psyllid impacts on Lupinus arboreus Sims, which consistently supported psyllid oviposition and development. The psyllid oviposited on all of the tested Lupinus spp., and completed two generations on five of the Lupinus spp., although numbers of psyllids were highest on French broom. In an additional experiment, A. hakani did not affect growth or survival of L. arboreus, but reduced growth and dramatically reduced survival of French broom. Taken together, these results indicate that Lupinus spp. are suboptimal hosts for the psyllid and are unlikely to be significantly impacted by its feeding.