Looking for a needle in a haystack – searching for hawkweeds

I’m in Falls Creek at the moment, with about 15 people from QAEG and other parts of Melbourne Uni. Cindy Hauser is running an experiment to detect hawkweeds, a major threat to the biodiversity in the Victorian Alps.

How would you look for a weed within the expanses of the Victorian Alps? Especially when it is hard to find when present?

Photo taken from Mt McKay by my son Owen – not a bad photo for a six year old.

Cindy Hauser and I solved this problem a few years ago, with the optimal surveillance strategy depending in part on the rate at which the target is detected when it is present. To estimate the detection rate, Cindy and her crew have planted hawkweeds (with locations plotted so that the plants can be removed again!) along with terrific replicas of hawkweed flowers.

A planted hawkweed with a replica orange hawkweed flower. Searchers looked for yellow versions that replicate the king devil hawkweed and rosettes without replica flowers (photo by Cindy Hauser; look closely and you can see snowflakes – it’s chilly up here!).

Surveyors have then been recording times to detections, from which the rate of detection can be estimated. That detection rate will then feed into the optimisation to help determine where the searchers should look, and how much time they should spend there.

The vegetation of the Bogong High Plains (which influences the detection rate) and the probability of hawkweed being present determine the allocation of surveillance effort that maximises the number of cells in which hawkweed is detected (from Hauser and McCarthy 2009). You can see Rocky Valley Storage in the vegetation map and the photo above.