Skip to content


  • After two years of research we have eliminated gregarine gut parasites from the tradescantia leaf beetle. This enabled Auckland Council to make the first releases of this agent against Tradescantia fluminensis, a high-profile weed that prevents indigenous forest regeneration.
  •  We developed a conceptual population dynamics model to help us understand the complex interactions between the nutrient content of heather and climatic factors causing the unexpectedly slow build-up of heather beetle in the central North Island. The model is also helping the Department of Conservation optimise future biocontrol of heather.
  •  We have conducted studies with Massey University that suggest that heather invasion in Tongariro National Park eliminates skink populations. We know of few examples globally where invasive weeds have such strong effects on indigenous vertebrates, although reptile biotas may have been understudied. Further research aims to understand the underlying causes, which could be due to weed-induced changes in habitat structure or in skink food resources.
  •  We have discovered that broom gall mites appear to use a symbiotic pathogen to create large galls, which can kill broom bushes. Alone, neither the pathogen nor the mites are effective. When considering gall-formers as biocontrol agents, we now need to consider possible roles of symbionts (either available in the New Zealand biota or as beneficial “associated organisms” in release applications).
  •  The Journal of Applied Ecology invited us to write a paper in their “Practitioners’ Perspective” series, on how applied ecological science could better assist weed biocontrol (submitted June 2011). This invitation from a prestigious journal demonstrates the high international reputation that the weed biocontrol team at Landcare Research has gained.


  •  In New Zealand, regional and national focussed end-users have developed weed distribution data-sets that are currently incompatible. We have succeeded in engaging the key players (DOC, MAF, Regional Councils and Future Forest Research) in a scoping study to develop an integrated, self-updating national weeds distribution database using modern software tools. This has the potential to improve weed risk assessment and weed management at regional and national scales.
  • We conducted the first quantitative analysis of global weed biocontrol success. This research highlighted several weed characteristics that influence biocontrol impact. Biological control is more effective against:  
    • Aquatic (vs. terrestrial) weeds
    • Biennial/perennial (vs. annual) weeds
    • Vegetatively reproducing weeds (vs. those only reproducing by seed)
    • Plants that are not reported as weeds in their native ranges (vs. those that are weedy everywhere)  

Managers can now, for the first time, rank their weeds quantitatively as good or difficult targets for biocontrol.

  • Our surveys of weed microbiotas, using plating and Denaturing Gradient Gel Electrophoresis DNA analyses, reveal unexpectedly high diversity of microbial endophytic species inside plants. This exciting discovery requires us to alter our view of weeds as single target organisms, and has considerable biocontrol (e.g. explaining unexpectedly poor agent performance) and biosecurity implications (e.g. new plant importations could unknowingly enrich endophyte biotas making existing weeds worse).
  • Our simulation models of Scotch broom populations show that the introduced seed beetle can cause local extinction of the weed but only if broom seed set is low. Incorporating the impact of varroa mite on feral honey bees provides just this scenario (as broom flowers need exotic bees to set seed). Thus varroa invasion may make broom less weedy by reducing honey bee pollination. Managers can assist broom control by excluding managed bee hives from areas where they may cause weed problems.