Assessing productivity of woodland birds

• It is possible to assess productivity of many woodland bird species by regularly surveying woods for fledged broods and comparing this with spring counts of adults.
• The technique could be used in studies of predation in woodland birds - we plan to undertake a study of the effects of grey squirrels on woodland birds in 2008/09.

The grey squirrel was introduced to England from America in the early part of the 20th century. It now occupies most of England, Wales and Ireland and is making inroads into Scotland and, on the continent, in Italy too. Wherever they go grey squirrels displace native red squirrels. Grey squirrels are much larger and live at higher densities than reds. They can kill trees by stripping bark so they are sometimes controlled by foresters.

There are also reports that grey squirrels predate birds, their nests and young. Some woodland bird species are declining and it is unclear why, so grey squirrels could be playing a role. If they are, they could be legally controlled wherever bird conservation demands it.

We are leading research to look at this and, with funding from the European Squirrel Initiative and Barnby Trust secured in 2007, we plan to compare the breeding success of birds in woods that have lots of squirrels, with woods in which squirrels are controlled for forestry or by gamekeepers.

However, there is no proven method for quantifying breeding success of woodland birds. Studies have monitored nests in trees to establish nest outcome but this is timeconsuming, expensive, and for species nesting in tree tops, almost impossible to do. For gamebirds and other ground-nesting birds, we assess breeding success using repeat counts of fledged broods. So we thought we should try this with woodland songbirds.

With the British Trust for Ornithology, we selected six 15-hectare woods and mapped bird territories to measure the total number of pairs in spring. We then surveyed the woods four or five times every week from May to August to look for broods.

Table 1 lists 19 bird species that we recorded at all six sites. Br:Pr is the brood-topair ratio for each of these. We had high encounter rates with broods of most species (indicated by Br:Pr approaching 1.0) on most days.

We can sub-sample our data to see how many times we actually need to visit each wood to calculate the Br:Pr ratios. For example, the next column is the probability of seeing any one brood if we surveyed woods just twice a week (2-visit prob). Broods of all but four species would have a >40% chance of being seen on this basis. Garden warbler is the common species for which this method is least likely to work.

Predation theory suggests that predators often take a surplus of a prey population so usually the proportion predated during breeding needs to be substantial if it is to affect adult population size the next spring. So we undertook a 'power analysis' which was designed to indicate how many sites we would need to study to detect a difference in Br:Pr of half of the mean value (the value in the table) for each species.

This means that a squirrel predation study is practical and we plan to apply this technique to a study in 2008 and 2009. The technique could prove useful for other studies involving productivity of woodland birds.