Search

Neodiplogrammus quadrivittatus (Oliver) and Trichapion lativentre (Beguin-Billecocq) are 2 South American weevils successfully used for the biological control of Sesbania punicea (Cavanille) Benth. in South Africa. The potential of these insects as control agents for the weeds Sesbania exaltata (Raf.) Rydb. and S. drummondii (Rydb.) Cory in the United States was studied in Argentina. Estimation of adult consumption, the food preference of adult insects on 5 species of Sesbania (3 native hosts and the 2 weeds mentioned above), the capacity of the insects to complete their life cycle on the weeds, and the weed infestation by wild populations in the laboratory garden were studied. The consumption of N. quadrivitattus on both north American weeds did not differ from that on native hosts. Similar consumption was observed for T. lativentre feeding on S. drummondii. S. exaltata showed the lowest consumption. The feeding preference of N. quadrivittatus was similar to that of its natural host plants. In the case of T. lativentre, results on only S. drummondii were comparable to that of its natural hosts. Both insects were able to complete their life cycles on the weeds. In the garden, there was a marked preference by the weevils for S. drummondii. S. exaltata was not attacked by T. lativentre and no test was carried out with N. quadrivittatus. We concluded that both insects are good biocontrol candidates for S. drummondii, whereas only N. quadrivittatus is a promising control agent for S. exaltata.

Sesbania punicea (Cav.) Benth. is a perennial legume that has become a weed of major importance in South Africa. Until recently, S. punicea was colonized by only one insect herbivore of any consequence, an introduced apionid weevil species, Trichapion lativentre (Beguin-Billecocq). The vegetative growth and reproductive capacity of S. punicea was shown to be significantly lower on plants at sites colonized by T. lativentre than on plants at sites without the weevils. However, these initial comparisons did not account for the extraneous effects of intersite differences, and additional analyses have been undertaken to determine the actual effects of the weevils on the weed. Regression models showed that between 77 and 83% of the variance in vegetative growth (leaflets, rachides, and stems) of S. punicea was explained by the presence of weevils, by their history of colonization, and to a lesser extent, by the density of the weevils, expressed as cumulative weevil-days. Kruskal-Wallis tests showed that the weevils also almost completely nullified the reproductive capacity of the plants by destroying nearly all of the flower buds. The results confirm that the suppression of plant growth was attributable, almost entirely, to weevil herbivory and that, in moderate to dense infestations of the weed, T. lativentre is equally effective as a biological control agent, regardless of site characteristics such as the spacing, height, or size of the trees.

Seeds of Sesbania punicea (Cav.) Benth. are largely impermeable, and the natural, low, percentage which germinate are permeable at the lens. Scarifying seed in acid stimulates germination and causes various types of damage to the testa, including the lens. Placing seeds in boiling water stimulates germination by rendering the seeds permeable at the lens. Seeds were mechanically scarified at precise sites on the testa to assess influence of the site and degree of damage on germination. Results indicate that both the site of initial water entry and the rate of water entry are important in determining seed and seedling vigour. Damaging the lens is more effective in stimulating germination and reducing seedling abnormalities than complete excision of the lens. This effect can be reversed if seeds from which the lens was excised are allowed to imbibe gradually in inert osmotica. The leguminous lens thus acts as the site of initial water entry and regulates the rate of water uptake, thereby increasing seed and seedling vigour.