The way we work

Spread across three sites in Canberra, Narrabri and Brisbane, we have developed facilities, systems and processes that can be applied to a wide range of agroecology projects. While we may lead projects across the Southern and Northern regions of Australia, we prefer to collaborate with our research partners in each state to improve the reach of our work.

Our day-to-day activities vary but include:

Defining the problem:  Our team actively  communicates and collaborates with our funders and stakeholders to ensure that we understand the scope of the problem in depth prior to commencing research. These discussions and activities allow us to understand the challenges faced by our stakeholders, and to determine how we plan the short and long term priorities of the the research.

Projecting the impact pathway: impact planning allows us to develop a scaled approach to our research, define the expected outputs and identify who we need to partner with to inform next users of outcomes and make recommendations.

Once a project has been developed we contribute in various ways.

On-farm surveys: much of our primary data for pests, natural enemies, pollinators and plants is collected by surveying across very wide geographic areas. We carry out once-off collection trips or regular trips involving a diversity of collection techniques including water traps, malaise traps, direct sampling off plants using leaf washes, suction samples,  sweep nets, beat boxes and beat sheets, pitfall traps, yellow sticky traps and occasionally, pheromone traps. Combining these techniques ensures a comprehensive sample of the communities and target species that occupy different niches in the crop. For example, a combination of visual counts (for Helicoverpa caterpillars and other Lepidoptera), leaf washes (for mites, aphids, thrips and whitefly nymphs) and suction samples (for flying pests such as mirids and beneficials such as ladybird beetles) covers assessment of  most of the pests and beneficials found in cotton. We are especially interested in the use of automated traps linked to image recognition to speed up the collection, sorting and diagnostic processes.

Controlled field experiments: by designing small scale on-farm trials and on research stations or with commercial collaborators, we minimise the effects of variables in our experiments. This helps us to discern the responses of invertebrate species to pesticides, assess the ability of plants to compensate for pest damage and to set better pest action thresholds. We also use this approach to determine how changes to the farming system (timing of planting, stubble load, crop type and variety) impact target species and populations and quantify the impact of pests on yield and crop production.

Automated trapping and identification: Our team is interested in findings new ways to quickly sample and identify the target species so we can generate large amounts of data about ecological processes. Some of these traps have been extended to surveillance tools for growers.

Experiments in controlled environments: if more stringent control of conditions is required to address our research question, we will experiment with species in shade houses, glasshouses and controlled temperature rooms. We are especially skilled at culturing a broad range of invertebrate species for experimental work involving live insects, spiders and mites, often on plants. Each species requires a different set of growth conditions and we have diverse mechanisms to control light, temperature and humidity to meet their unique needs.

Data analysis and synthesis: A large and important part of our research role is the quality control, analysis and synthesis of the data we collect or use for addressing research questions. Taking complex and large data sets and extracting information that allows us to answer a research question is critically importantA recent example of such an analysis that has demonstrated better outcomes for pest pressure, insecticide use, and yield in less intensive agricultural landscapes, was derived from data available in crop consultant monitoring reports. 

Gagic, V., Holding M., Venebles, W, Hulthen, A and Schellhorn N. (2021) Better outcomes for pest pressure, insecticide use, and yield in less intensive agricultural landscapes. Proceedings of the National Academy of Sciences. 118 (12) e2018100118; DOI: 10.1073/pnas.2018100118

Large fields embedded in simple landscapes have the earliest pest immigration, highest probability of pest presence, and lowest yield. Model predictions for the (A) pest immigration (time elapsed from crop planting to the first pest recorded), (B) probability of pest presence in unsprayed fields, and (C) yield in relationship to field sizes and proportion of seminatural areas.


Predictive computer simulation models: Simulation models can be used to gain further insights into not only the drivers of agro-ecological system dynamics but also enable us to forecast and predict outcomes of varying landscape management scenarios. For example, Hazel Parry has worked for over ten years in the team using this approach to address a wide range of questions in multiple agricultural systems both here in Australia, and recently overseas studying whitefly infestation of cassava in Africa to evaluate cultural control and resistance-breeding strategies for pest suppression. 

Molecular tools: The use of  molecular tools is an advanced,  and now common, methodology for the identification of target species. We partner with molecular ecologists in CSIRO, collaborating with Wee Tek Tay’s and Tom Walsh’s laboratory, and other research organisations (University of Melbourne, Hoffmann group) to deliver samples, process the samples so they are ready for DNA extraction, at times helping with the initial molecular steps.

Cassava whitefly pests, Bemisia tabaci are often challenging to tell apart morphologically.

Cassava whitefly pests, Bemisia tabaci are often challenging to tell apart morphologically. This one is likely to be the SSA1 type but only molecular techniques can verify this. Nymphs and pupae feed and develop on the underside of a cassava leaf.


Diversified use of data sets: we are open to sharing data and are always interested in hearing from people who may like to use our data sets for new purposes and to gain new insights. Some examples of our data sets can be downloaded from the CSIRO data access portal.

  • Parry, Hazel; Macfadyen, Sarina; Holloway, Joanne; Severtson, Dustin; Hoffmann, Ary; Umina, Paul; van Helden, Maarten; Binns, Matthew (2020): New Knowledge on Pests and Beneficials in Grains. v1. CSIRO. Data Collection.
  • Hill, Matt; Umina, Paul; Macfadyen, Sarina (2018): Imidacloprid seed-dressings and beneficial invertebrate communities in broad-acre grains production. v1. CSIRO. Data Collection.

Communicating our findings: an important part of our work is to convey our findings through various media  and networks. This includes digital tools, fact sheets, industry articles, field days, training and best management guidelines, and scientific papers.

Capacity building: a core output for every project we work on is to build capacity for Australian scientists and overseas research teams. This may involve peer-to-peer learning on specific tasks like developing online forms for use during a field survey, data analysis using R or small group training such as the  identification of certain invertebrate groups. We are constantly learning new and better ways of doing our research and are keen to pass these skills onto others if requested.

The photos below were taken from a recent training course on the identification of parasitoid wasps, which are natural natural enemies of cassava whitefly pests, held in Uganda, it was run by Dr Wee Tek Tay (CSIRO) and Dr Andrew Polaszek (NHM, London) as part of an international project led by the Natural Resources Institute, London.