The researchers looked at 90 woodland trails a few kilometres north of Sydney, half regularly used by dog-walkers and half where the animals were prohibited.
Dogs were walked, on leads, along the 250m-long (820ft) trails, followed 20 seconds later by an observer who counted the birds seen and heard. The experiment was repeated for walkers without dogs and for a control scenario where there were neither walkers nor dogs.
The team found that where dogs were walked saw 41% lower numbers and 35% fewer species. Results were similar in areas normally used by dog walkers suggesting that birds don't get used to dogs.
Humans walking without dogs caused a reduction in numbers but generally less than half that caused when dogs were walked.
Team member Dr Banks said:
The birds were clearly showing an aversion to dogs - they clearly perceived dogs as a potential predator.
Martin Fowlie, a spokesman for Birdlife International, said the study confirmed what conservationists already suspected.
This is not a surprising result; there is already evidence that dogs can disturb birdlife, but it is interesting that someone has now quantified it.
"It would now be really interesting to see how long those reductions in bird numbers last, to see whether it is a few minutes, hours or days.
Would doggy fans note that this is a report and please don't shoot the messenger. Hello Jan ;)