A method for mist-netting breeding Eurasian Woodcock: use of visual and audio lures increases capture rate

ABSTRACT

We describe a method for mist-netting Eurasian Woodcock Scolopax rusticola in the breeding season using a remote-control playback lure and a decoy. Nineteen roding woodcock were caught in 39 sessions in 2016. A GLM, in which length of mist-net was specified as an offset, was used to compare our capture rate to that of a previous study in which no lure was used, suggesting our ‘number of captures per session’ was approximately nine times higher. Ringing individuals that are known to belong to Britain’s resident breeding population could provide more comprehensive data, which are of particular value given this population’s ongoing decline.

Ringing has provided a means of studying chick survival, winter dispersal and philopatry in the Eurasian Woodcock Scolopax rusticola as well as a rudimentary understanding of the species’ migration (Hoodless & Coulson 1994, 1998). The number of Woodcock ringed in Britain & Ireland has increased dramatically in recent years, with an average of 65 ringed per year during 2000–05 rising to over 1 300 individuals each year during 2010–15 (Robinson et al 2016). This is largely due to the increased use of dazzling to catch Woodcock following a concerted effort to promote the method (Williams 2015). Dazzling is conducted at night during the winter months, when Woodcock leave woodland to feed on arable fields and grassland.

Woodcock caught in winter may belong to Britain’s small breeding population (c55 240 males; Heward et al 2015) or to one of the larger Russian, Baltic or Fennoscandian populations which overwinter in western Europe (c800 000 to 1.5 million individuals winter in Britain each year; Musgrove et al 2011). Trends differ between these populations, with most continental migrant populations appearing stable (BirdLife International 2015, Lindström et al 2015) compared to the marked decline of British residents (Heward et al 2015). These populations cannot be separated morphometrically. Lower Woodcock densities and seasonal changes in their nocturnal behaviour make dazzling unproductive during spring and summer (Hoodless & Hirons 2007), but an alternative method for capturing Woodcock during the breeding season, when only residents are present, would be beneficial. An approach similar to the BTO’s Retrapping Adults for Survival (RAS) scheme, at a variety of sites, would be a useful start towards an improved understanding of the demographic drivers of decline, provided that sufficient birds could be captured.

A number of studies have used mist-nets to catch Woodcock (Sheldon 1957, Hirons 1980, McAuley & Longcore 1993, Hoodless & Hirons 2007), although no comprehensive guide to this catching method has been published. Literature regarding catching techniques for American Woodcock S. minor during the breeding season is not directly applicable to the Eurasian Woodcock, given the differences in their breeding displays (McAuley & Longcore 1993). Older studies of Eurasian Woodcock have relied on the use of a live bantam decoy, as described by Hirons (1980). There are animal-welfare considerations associated with this method and, in the UK, a live decoy cannot be used without the prior consent of the ringing unit (Redfern & Clark 2001). Some researchers have caught small numbers of Woodcock in spring and summer using walk-in traps in woodland (Hoodless & Hirons 2007) or by locating incubating females or broods using pointing dogs (McAuley & Longcore 1993). Here, we summarise an efficient technique for mist-netting Woodcock in the breeding season.

During this study, in 2015 and 2016, we captured Woodcock in three large (>100 ha) woodlands in Nottinghamshire, UK, managed by the Forestry Commission. Mist-netting was conducted at dusk, from mid-April to mid-July, when male Woodcock perform their breeding displays (known as roding). Capture sites were usually selected based on preliminary counts of roding Woodcock and were located in clearings and wide rides. Catching was most successful at sites where ground vegetation was low or patchy. Catching sites included heaths, mown areas such as car parks and picnic sites, clearfells, permanent clearings covered with low vegetation or dead bracken and woodland rides. The sizes of clearings used were 0.5–4 ha.

Mist-nets were erected 30  min before sunset and sessions were run until 60–90  min after sunset. We used Ecotone 45x45-mm mesh nylon nets (110 denier/2 ply) with four shelves. Up to four nets were used, with a maximum combined length of 60 m (two 12-m, two 18-m). Nets were most commonly arranged as a four-sided ‘box’ or an open three-sided one, often using a woodland edge to form the fourth side. Occasionally nets were set in a continuous line dog-legged across the clearing. On forest tracks and rides, we set three separate nets perpendicular to the ride.

Woodcock were attracted to nets using a playback lure and a decoy. The playback lure was placed near to the centre of the box of nets or, where parallel rows were used, to one side of the central net. Three types of decoy have been used, each separately: a taxidermied Woodcock, a papier-mâché model with a real Woodcock tail and a motorised ‘weasel’ cat toy, the latter providing a moving stimulus. The importance and relative effectiveness of these three decoys have not yet been quantified. An observer hid in nearby vegetation, usually within 8–15 m of the nets, and used a remote control to select between different calls stored on the playback lure. Any UK ringer wishing to use playback lures during the breeding season will need a specific endorsement on their ringing permit.

When a roding Woodcock was seen or heard it was drawn close to the nets using the high-frequency ‘whistle’ element of the roding call (a recording with the low-frequency ‘grunt’ component removed). When a bird was seen to alter its flight towards the call, the playback was switched to a ‘chasing’ contact call. This chasing vocalisation, a rapid series of whistles, is usually used during male–male confrontation and can elicit a noticeable change in behaviour. The playback lure was most effective when its use was restricted to the second half of the roding period, generally between 30 and 60 minutes after sunset. Nets were watched by the lure operator whenever the playback was in use and birds were extracted from the nets as soon as they were caught. When nets were set in a box arrangement, operators positioned themselves so that any Woodcock landing in the centre of the box without becoming caught might be flushed towards the nets.

We compared our capture rate using this new technique with that recorded by Hoodless & Hirons (2007) at Whitwell Wood, Derbyshire, during April–June 1991–92. At Whitwell Wood, Woodcock were captured in mist-nets without use of playback or a decoy. Whitwell Wood lies within 15 km of our current study sites and the number of registrations of roding Woodcock in 1991–92 was similar to recent counts at our current sites (15–25 per evening roding period). Typically, 126 m of mist-net (one 18-m and nine 12-m nets) were opened for 75–90  min, with 51 dawn sessions and 49 dusk sessions over the two years. Nets were North Ronaldsay 32x32-mm mesh (75 denier/3 ply), except one 12-m net with 38x38-mm mesh. Observations of Woodcock hitting various Ecotone and North Ronaldsay mist-nets with mesh sizes 30–45 mm suggest similar likelihoods of capture; certainly Woodcock were never observed to bounce out of any of these mist-nets. At smaller mesh sizes (eg 16–19-mm square mesh), however, this is more likely to occur.

Capture rate was analysed using a generalised linear model (GLM) with captures per session as the dependent variable, a Poisson error distribution and log_e(length of net) as an offset. Factors were the combination of method and time of day (net only at dawn; net only at dusk; net and lure at dusk) and time of year (summarised as two-week periods corresponding to the first and second half of each month).

We caught 15 Woodcock in 2015 and 19 new Woodcock in 2016. In 2015 we recaptured three birds. In 2016, we made one same-year recapture and recaptured three birds originally caught in 2015. We recorded no mortalities and no injuries to Woodcock as a result of mist-nets or during handling. After experimenting with the technique in 2015, the method outlined above was consistently repeated for 39 sessions in 2016, with birds caught between 19 April and 13 July. Birds were captured, on average, 54 ± 19  min (1 sd) after sunset. Of the 22 Woodcock caught in 2016 (19 new and three recaptures), 19 (86%) were confirmed as male. Fifteen were sexed based on display behaviour immediately prior to capture. The remaining four birds were sexed based upon their ratio of bill to tail length (about 45% of adult birds can be sexed in this way; Ferrand & Gossmann 2009) or their subsequent behaviour (Woodcock were fitted with radio-tags). Two birds were unsexed and one was confirmed as female by the remnants of a brood patch. A strong sex bias is unsurprising given that the playback method uses a recording of a vocalisation usually used between males.

There was no difference in capture rate with mist-nets between 1991 and 1992, so data were pooled for further analysis (GLM year F_1,97 = 0.10, P = 0.753, time of day F_1,97 = 8.77, P = 0.004, year x time of day F_1,96 = 0.55, P = 0.462). There was a consistent difference in capture rate between methods, and between dawn and dusk sessions, during April to mid-July (GLM method-time F_2,130 = 22.18, P < 0.001, two-week period F_6,130 = 1.25, P = 0.285, method-time x period F_9,121 = 1.46, P = 0.169, Figure 1). At Whitwell, capture rate was almost three times higher at dawn than dusk when using mist-nets alone (mist-net only at dawn 0.157 ± 0.030 capture/50 m of net; mist-net only at dusk 0.055 ± 0.018). Comparing between methods, capture rate at dusk was over nine times higher when using the playback lure (playback lure 0.527 ± 0.098). Results based on captures per session, ignoring the difference in number of nets used between methods, were similar (GLM method-time F_2,130 = 8.60, P < 0.001, two-week period F_6,130 = 1.37, P = 0.231, method-time x period F_9,121 = 1.58, P = 0.128; playback lure 0.590 ± 0.110 capture/session; mist-net only at dawn 0.412 ± 0.080; mist-net only at dusk 0.143 ± 0.047).

Figure 1. Comparison of capture rates of Eurasian Woodcock with mist-nets and a playback lure at dusk relative to mist-nets alone at dawn and dusk. Error bars are ±1 standard error.

Our new method of capturing Woodcock in the breeding season appears to be more efficient than mist-netting alone and requires fewer nets. Because comparisons are made between two different studies, conducted at different sites and times, it is possible that these confounding factors partially account for some of the observed difference in capture rates. However, given the ninefold increase in capture rate associated with the use of our remote-control lure, it seems unlikely that any more than a small proportion of this difference is attributable to site and timing effects. It might be possible to improve the capture rate further by employing the technique at dawn.

The capture method described here is not without its limitations, chiefly the skewed sex ratio of birds caught, with females poorly represented. Nevertheless, ringing Woodcock during the breeding season using this method over several years could generate valuable data on annual survival rates of resident breeders and help in determining the drivers of their decline, provided that any potential sex-specific effects are considered. The combination of mist-netting during the breeding season and dazzling on nearby fields in winter could be used to provide additional recaptures and perhaps also seasonal estimates of survival. Monitoring overwinter survival at a sample of sites would be valuable, given that the impacts of hunting and harsh winter weather are poorly understood.

Acknowledgements

We are grateful to the Forestry Commission, Nottinghamshire Wildlife Trust and other landowners that provided access to their land. We thank the members of Birklands Ringing Group for assistance in the field, particularly Ann Ward. We are grateful to several researchers for discussions on catching Woodcock and ANH is indebted to Vadim Vysotsky and Mikhail Verevkin for demonstrating a similar playback technique used in Russia. We thank Juan Arizaga and an anonymous reviewer for comments made on an earlier draft.