Understanding the biology of tropical soda apple (TSA) is necessary to develop effective control strategies. Tropical soda apple seedling emergence, growth, and regeneration from roots were evaluated. Plant growth (height, leaf number, flowering) was monitored over 110 d in a greenhouse experiment. Emerged seedlings required 83 d to reach a height of 40 cm and 108 d to flower. Total nonstructural carbohydrate concentrations in TSA root and stem fractions, monitored monthly for 1 yr, were highest in December (35%-root, 18%-stem). Seedling emergence as affected by planting depth (0, 1, 2, 4, 8 cm) and regeneration from root segments (7.5, 15 cm long) placed on the soil surface and 5 and 10 cm deep were studied in separate greenhouse experiments. Seedling emergence was lowest (P < 0.05) for seed placed on the soil surface (6.9%) and at a depth of 8 cm (19.4%) with the highest level of TSA emergence predicted at 63% for seed planted 3.6 cm deep. Regeneration from roots was higher (P < 0.05) for the 15-cm root segments than for the 7.5-cm segments while depth of root segment placement did not affect regeneration. Weed control strategies should involve cultural and chemical practices that affect both roots and seeds.

Tropical soda apple is an aggressive prickly perennial shrub growing up to 2 m high. It invades open to semishaded areas, including pastures, forests, riparian zones, roadsides, recreational areas, and horticultural and cropping areas. In Australia during August 2010, the weed was identified on the New South Wales Mid North Coast. It is believed that tropical soda apple has been present in this area for a number of years and both systematic and ad hoc surveys have found the weed in other satellite locations. The discovery of tropical soda apple at several cattle handling facilities indicated that cattle are a significant vector for the weed. The aim of this project was to use the National Livestock Identification System (NLIS) data to trace cattle movements from affected properties throughout New South Wales and into other Australian states. This has proved advantageous, as there are few other nonecological mechanisms to systematically trace significant weed movement. We have been able to conduct a pathway analysis of where this weed is likely to occur across New South Wales through the use of NLIS. Importantly, we can use this information to pinpoint surveillance activities for local managers, thus ensuring better use of resources. We have also been able to create a stochastic model for incursions at these sites using information gleaned from the NLIS data.

Tropical soda apple samples were collected from 31 populations across the southeastern United States and from four populations in a portion of its native range in Brazil. The genetic relationships among these populations were examined by single primer amplification reactions (SPAR) and by sequencing a portion of the chloroplast genome. SPAR revealed no variation among the 132 individuals and only two chloroplast haplotypes were detected in the 50 individuals sequenced. The most common haplotype was present in all samples from the United States and most of the Brazilian samples, whereas the second haplotype was only found in one of the Brazilian populations. Within the limitations of these data, we conclude that Brazil is the best location to seek a potential biological control agent for tropical soda apple, and that, if identified, this agent should prove useful for populations throughout the United States.