How many breeding woodcock?

The breeding woodcock population in Britain was estimated by a new, species specific method at 78,000 males, significantly greater than the 5,000-12,500 pairs estimated by general bird surveys.  There were substantial differences in densities between regions and woodland types.

Owing to its cryptic plumage, secretive behaviour and nocturnal habits, the woodcock is a difficult species to survey. The breeding population size is currently estimated at 5,000-12,500 ‘pairs’, based upon sightings made during the course of general bird surveys. Counts of passes by roding woodcock (males in display flight) provide the only feasible method for any dedicated large-scale survey, but their interpretation has been hindered by the fact that they represent multiple registrations of an unknown number of males. Recently, we have demonstrated a relationship between counts of passes and numbers of males and hence the validity of roding woodcock counts for population monitoring (see Review of 2003). Here we update the preliminary results from our 2003 breeding survey (see Review of 2004), providing revised population estimates and examining woodcock densities in further detail.

The survey involved volunteer observers making three counts of woodcock passes in the largest wood within 807 randomly selected one-kilometre squares that were stratified by 11 regions and four wood size classes. The number of individual male woodcock at each survey site was estimated from the maximum number of registrations using our calibration equation. The estimated number of males at each site was assumed to be equivalent to the density in the one-kilometre survey square, because the average roding area of woodcock is known to be 88 hectares.

Woodcock presence varied between regions and wood size classes, with higher occurrence in larger woods. Breeding woodcock were more widely distributed in Scotland and northern England than in southern England and Wales. Weighting by the availability of one-kilometre squares within each region-wood size class gave a national estimate of 35% presence in squares containing at least 10 hectares of woodland. Average woodcock density in occupied woods was 2.76 males per 100 hectares, but there was large regional variation, ranging from 0.87 males per 100 hectares in Wales to 4.10 males per 100 hectares in East Anglia. Perhaps not surprisingly given that we employed a dedicated survey method, our estimate of the national woodcock population at 78,000 males (95% confidence interval 62,000-96,000) is far higher than the general bird survey estimate. Our survey showed that Scotland (39,000 males, 95% confidence interval 24,000-57,000) and England (37,000 males, 95% confidence interval 30,000-44,000) support similar numbers of woodcock, with only 2,000 males (95% confidence interval 1,000-3,000) in Wales.

Figure 1. Male woodcock density in relation to woodland stand type

Several aggregations of breeding woodcock were apparent in large forests, such as Kielder Forest, Dalby and Newtondale Forests, Thetford Forest, Forest of Dean and New Forest, and heavily wooded regions, such as Derbyshire and Nottinghamshire, West Sussex and north Hampshire. These appear to be important strongholds for breeding woodcock and ensuring appropriate management of these forests will be an important step towards securing breeding woodcock populations in the future.

Our overall estimate of 35% woodcock presence suggests that there remains much potential woodland habitat that currently supports no woodcock. Absences from woods larger than 50 hectares, in particular, suggest unsuitable habitat structure or local population decline. Woodland occupancy in Britain is comparable to that in France (20-30%) and Switzerland (19-31%), but in Russia, which is believed to be the main stronghold for the species in Europe, woodcock are present in 85-95% of forests.

By classifying records of dominant trees and vegetation at count sites into stand types, we were able to examine woodcock densities in relation to different habitats. This revealed that, in conifer forests, stands of Scots/Corsican pine supported higher densities than those of Douglas fir/larches or Norway/Sitka spruce, whereas among deciduous woods, alder/willow had higher densities than oak/ash or beech (see Figure 1). Overall alder/willow woodland was the best habitat, supporting woodcock densities eight times higher than in spruce, the worst habitat. It should be borne in mind that this information relates to displaying males and does not necessarily reflect where females choose to nest, although males are thought to rode most intensively over the best nesting habitats. These results are in broad agreement with our earlier intensive studies of radio-tagged woodcock, which showed that stand use was related to habitat structure and earthworm availability.