Ross Carter-Brown is a postgraduate student in the Department of Ecology at Lincoln University. Here he writes about an ongoing project that started off as a summer scholarship.

Restoration projects tend t  o focus on mass plantings to revegetate a site and achieve canopy closure as soon as possible. These plantings usually consist of an initial planting of a relatively small number of hardy species, followed by a secondary planting of less hardy species years later to inject additional diversity. In recent years many ecologists have thought that expediting the return of ecosystem services, such as seed dispersal by birds, may aid the revegetation process and the restoration process overall. One way of accelerating seed dispersal is to provide perching structures in the restoration site to encourage visiting birds to make a deposit (including the seeds in their poo!).

Bird perches have been tested extensively internationally with mixed results. Many studies  only used one site, and focused solely on the plant species and the number of seeds that birds were depositing. Simply counting and identifying the seeds that arrive under bird perches is only part of the puzzle. Those seeds must then survive roving seed predators, such as mice and rats, long enough to germinate. Small seedlings with tiny roots systems must then weather highly variable soil moisture levels as well as herbivores before they can become big enough to thrive. We set out to test the effectiveness of artificial bird perches at increasing seedling recruitment at restoration sites. We measured the amount of seeds being brought in by birds as well as seed predation rates and the all-important seedling establishment rates. A host of location variables were also recorded.

Sticking a big bird perch in the middle of a paddock seemed like a bit of a long-shot to us, let alone knowing if any birds would actually use a big ungainly perch far from the relative safety of the bush. Would any seed deposited by the birds stand a chance of germinating and growing amongst a thick sward of pasture grasses?  We decided that we would stack the odds in our favour. We opted to erect our perches in established restoration plantings where much of the grass has been shaded-out through recent canopy closure. On top of this, we also decided to bribe the birds with fresh banana at the top of the perches on a regular basis. Three sites were selected for the experiment. Quail Island in Lyttleton Harbour (Canterbury), Kate Valley in North Canterbury, and The Punakaiki Coastal Restoration Project (just south of Punakaiki) on the West Coast, all in the South Island. All of these sites had quite unique characteristics which we hoped would provide us with differing results. In order to measure seed predation rates we also set-up a parallel experiment at the same locations as the perches. This involved placing out seeds of a number of native species with different seed sizes. They were placed in a series of small plastic rings on the ground, with half of the replicates having exclusion cages over them as the control. The cages excluded mammals and birds but as I found out, some invertebrates still showed up for lunch!

Although we are yet to process all the samples we still have some idea of what has been happening from our observations. We have witnessed a large number of silvereyes and a small number bellbirds using the perches and eating the bait at Quail Island and Kate Valley, but only while the bait is present. We can also see very clear differences in the amount of seed being collected in seed traps between the perch and control plots. Punakaiki is a completely different story. Over the last 12 months we didn’t observe any birds landing on any of the perches or eating the bait despite both silvereyes and bellbirds being present. This was backed up by the observations of the Conservation Volunteers (CVNZ), who were onsite five days a week and there was no difference in seed deposition between the perch and control plots. Why are we getting such different results between the sites? At this stage we really don’t know.

The 12 month experiment is now drawing to a close with only one final trip to the Kate Valley site to be completed. Over the last year we have collected over 1000 trap samples and these are yet to be processed and analysed, so there is still much work to be done. If you have a yearning to learn how to identify a range of native seeds, please get in touch. I have lots for you to look at! Check back in few months when I’ll have some more answers for you.

This research would not have been possible without the generous funding by Brian Mason Scientific and Technical Trust, Rio Tinto and Lincoln University. There has also been many hundreds of hours of volunteer labour provided by family, friends, and colleagues who I am indebted to.