1 Landcare Research New Zealand, PO Box 40, Lincoln 7640, New Zealand
2 Landcare Research, 231 Morrin Road, Auckland, 1072, New Zealand
3 Universidade Federal do Paranê, Rua Bom Jesus, 650 Juvevê, 80.035-010 Curitiba, Paraná, Brasil
4 BioDiscovery New Zealand, 24 Balfour Road, Parnell, Auckland, New Zealand
Tradescantia fluminensis (Commelinaceae) was introduced into New Zealand (NZ) as a house plant, but is now a serious under-storey weed of indigenous forest. Surveys for potential biocontrol agents in SE Brazil, starting in 2005, identified a rich natural enemy biota including herbivorous insects and plant pathogens. Routine screening of the first insect agent to be host range tested, the leaf beetle Neolema ogloblini (Chrysomelidae), revealed high levels of a gregarine (sporozoan protozoan) gut parasite. This appeared to reduce beetle fecundity, longevity and general vigour, potentially compromising its biocontrol efficacy. Depending on the host specificity of the gregarine, it could also threaten NZ fauna. In NZ biocontrol agents can only be released from containment if they are shown to be free from unwanted associated organisms. We report on two years of increasingly intensive attempts to obtain a gregarine-free population of N. ogloblini including use of highly hygienic field collection methods in Brazil to get clean material at source, surface sterilisation of eggs, use of cages with HEPA-filtered air in containment, and attempts to improve our gregarine detection methods by gut dissection and DNA probes (both of which proved less easy, and more expensive than anticipated). In December 2010 we finally released N. ogloblini from containment after showing we had three consecutive generations of beetles tested negative for gregarines. Success was achieved by repeated sub-culturing. Firstly, eggs were collected as hygienically as possible from single female beetles (each having been paired with a single male). Then each larva was reared in solitary containment but with poor hygiene to ensure that any low level of gregarine infection would be expressed sufficient to minimise the risk of getting false negatives in subsequent testing. All lines testing positive were eliminated. Final crossing of lines before release from containment was carried out in an attempt to restore lost heterozygosity and overcome any inbreeding depression or adaptation to laboratory conditions.