Department of Statistics


Case study: Helping Tiritiri Matangi flourish

Hauraki Gulf wildlife sanctuary Tiritiri Matangi Island is one of New Zealand’s most important reforestation projects.

cp-chris-triggs-pohutukawa

Since the 1980s, more than 280,000 native tree seedlings have been planted, around 90,000 of them pohutukawa, to help boost populations of native birds, plants, and invertebrates like weta and native snails.

However, the island ended up with a problem: the pohutukawa canopy in replanted areas became so thick that little light and few seeds penetrated, preventing animal and plant life from thriving.

A possible solution: Cut light wells into the canopies. But a careful experiment needed to be done to find out the best way to do this.

Statistical Consulting Centre head Chris Triggs, right, is a long-time Tiri volunteer and donated his time to the project. He helped clarify the goals of the research and designed a robust data collection process.

Five areas of pohutukawa monoculture were designated for study, with each area split into halves. Treatments were randomly assigned to each half: either single trees were removed to create small light wells; all trees in a 15 sq m area were removed to create larger wells; or nothing was done, to create a control area.

As well as designing a statistically-robust sampling scheme in order to be able to draw valid scientific conclusions, says Chris, the areas under study needed to be accessible and safe as Tiri’s army of keen volunteers would be doing the data collection – counting insects in traps, birds in the trees, and plants underfoot.

The first light wells were cut in April and May 2011. Monitoring will be carried out twice a year, and Chris will be analysing the data to find out what is happening to bird, insect and plant populations in the areas under treatment.

His statistical work will make good use of R, the statistical computing and graphics software developed at the Department of Statistics in the mid-1990s and now used all over the world. “We’re an R shop,” says Chris, “and this sort of work is ideal for it.”