The full paper is published in Henderson et al., 2017. Population estimates and habitat associations for Stonechat (Saxicola rubicola), Whinchat (Saxicola rubetra) and Wheatear (Oenanthe oenanthe) in Wales, 2012 and 2013. Birds in Wales.
In this survey, we generated some basic associations between species and habitats or landscape characteristics. Relationships for Wales are consistent with our general knowledge of the species’ natural history, but components underpin the growing consensus that land management has contributed to a squeeze on niche availability for Whinchat in particular (Calladine and Bray 2014). Conservation wise, Whinchat deserves the most urgent attention due to a rapid population decline, even in this apparent stronghold in Wales. Key characteristics of breeding habitat were consistent with the grassland and vegetation structural variation found elsewhere. Whinchats require suitable perches from which to forage or display (e.g. tall flower stems, bracken, light scrub or small trees), by short swards, and tussocks in which to nest safely and forage (Border et al. 2016, Murray et al. 2016). Such qualities are now mostly retained within marginal and transitional habitats, such as wet areas or ffridd, that exist due to circumstances of topography, water logging or fencing (around young plantations), to limit habitat homogenisation through grazing or management (Fuller et al. 2006). An association with damp features for Whinchat is consistent with this species’ tendency to select an insect rich, temperate micro-climate in Europe, including lowland wet meadows in Slovenia (Tome and Denac 2012), meadow slopes in Switzerland (Müller et al. 2005) and lower grassy valley slopes on Salisbury Plain, that are wet in winter but lush in summer (Border et al. 2016). In upland boggy habitats, drying, through drainage management and climate warming, can significantly reduce the abundance of important prey species, such as crane flies, for species such as Whinchat (Carroll et al. 2011, Pearce-Higgins 2011).
Stonechats by comparison, are short distance migrants that in Europe can thrive in drier breeding conditions than Whinchats, including Mediterranean habitats and lowland heathlands where breeding Whinchats are rare (Hagameir and Blair 1997). The species appears to be better adapted to benefit from relaxed thermal limitations, which improves survival rates and opens up new habitat opportunities (Jiguet et al. 2009). This has undoubtedly contributed to range expansion in Wales and the UK generally. Whinchats, by comparison, with a relatively short breeding window, rely on predictable, high-season prey abundance. As long distance migrants, they are more susceptible than Stonechats to a disruption in the timing of peak prey availability (Visser et al. 2015). These contrasting ecological requirements between the two species offer a likely explanation for their diverging population trends (Juiget et al. 2007), via differences in survival rates and/or productivity.
The survey also allowed population estimates to be formally estimated for Wales for the first time, and in spite of wide margins of error, population estimates are calculated using a standardised survey protocol that is defined and repeatable. Earlier estimates appear to be based on densities or occupancy rates calculated at varying scales, so the decisions used for extrapolation are unclear, but probably involve expert judgement. This means current and past estimates are difficult to compare.
For Whinchat, Lovegrove et al. (1994) estimated 5,000-6,000 pairs for 1994. The BBS trend for 1995-2012 was -53%, which following Lovegrove et al (1994) would suggest around 2,500-3,000 pairs for 2012. Independently, the British population estimate was 46,864 in 2009 (range: 18,689-75,000 pairs) also based on the BBS (Musgrove et al. 2013), and the Welsh proportion of the atlas (Balmer et al. 2013) for Whinchat was around 10% in terms of occupied squares, or 4,900 pairs. However,the British estimate was simply calculated from the number of individuals, halved which potentially overestimates true territory occupancy. Since 2009, the rate of decline in Whinchats in Britain has slowed (Harris et al. 2016), therefore, again, our current mid-estimate for Wales of perhaps 3,000 pairs seems reasonable. Against this rationale, the Green (2002) estimate of 2,400 pairs for year 2000 appears low.
For Stonechat there have been rapid, weather-related, fluctuations in population status, though principally increasing by around 250% between 1995 and 2012 (Risely et al. 2012). Detectability is far less of an issue for this species than for Whinchat, and it is more likely to occur in widely scattered territories at low density, potentially on sub-optimal land. Our estimate using representation across landscape types lies between 9,736 and 14,448 pairs (95% confidence range: 7,342–22,925). Green (2002) suggested 3,000-3,500 pairs (year 2000), based on the national atlas survey 10 years earlier (Gibbons et al. 1993), though in fact the species was increasing in the intervening years, and indeed the BBS trend for 1995-2014 was a significant 93% (Harris et al. 2016), almost double. The BBS may also under-represent change in remote regions of Wales where the species has increased most (Balmer et al. 2013). The estimate for Britain in 2009 was 56,000 pairs (Musgrove et al. 2013), with the Welsh contribution being around 20% of occupied squares (Newson pers. comm.) suggesting 11,200 pairs. The 9,000 to 14,000 pairs range for Wales therefore seems reasonable.
Wheatear was difficult to survey accurately using a multi-species, approach, in terms of survey timing and not knowing where to ‘draw the line’ in terms of defining passage versus breeding status. Wheatear has a protracted migration, which increases the risk of breeding status being assigned to birds that are still on passage, even in May, and particularly on the coast. Lovegrove et al. (1994) and Green (2002) provided very different estimates of 3,500-4,500 pairs and 5,000-10,000 pairs respectively but with no clear or comparable rationale. Against these values, our lower estimate (13,759 pairs – 95% confidence range: 7,153–25,861) appears high, despite being calculated by excluding April observations and larger coastal groups. The estimate for Britain in 2009 was 230,000 pairs (Musgrove et al. 2013), with the Welsh contribution being around 5% of occupied squares (Newson pers. comm.) at around 11,500 pairs. Brenchley et al. (2013) found a marginal increase of 9% in the number of occupied 10-km squares in North Wales between 1991 and 2008-12, while the UK atlas (Balmer et al. 2013) calculated a 10% decline in occupied squares in Wales generally over that same period, suggesting there had been shifts in distribution but no widespread population increase. However, for this species in particular, without a standardised, structured and representative sampling framework, and critically, a defined and consistently applied rationale for identifying ‘pairs’ and a method for representing sub-optimal habitats properly, both absolute population estimates and relative comparisons between surveys will be especially difficult to make with confidence. This species may require a dedicated survey method for accurate assessments of population size to be made
- Final coverage was just short of 300 different 1-km squares with chats recorded in 63.4% of visited squares.
- All species were associated with extensively managed, non-intensive, unimproved, habitats, and none with linear features such as hedges.
- Both Stonechats and Whinchats forage from prominent features and were associated with scrub or bracken but Stonechats were also associated with coarser herbaceous cover and Whinchats with semi-natural grassland and damp areas.
- In many places, semi-natural grassland may only be available in areas protected from heavy grazing, such as fenced young tree plantations or boulder strewn steep slopes where rank herbaceous cover is also less likely to dominate.
- Due to land management and grazing, the natural grassland components of the habitat favoured by Whinchats especially will continue to become increasingly dispersed if protected only by topography (inaccessibility) or incidental exclusion (fenced plantations).
Gibbons, D.W., Reid, J.B. and Chapman, R.A. 1993. The New Atlas of Breeding Birds in Britain and Ireland: 1988–1991. London: T. & A.D. Poyser.
Green. J. 2002. Birds in Wales 1992-2000. Blaenporth: Welsh Ornithological Society.
Harris, S.J., Massimino, D., Newson, S.E., Eaton, M.A., Marchant, J.H., Balmer, D.E., Noble, D.G., Gillings, S., Procter, D. and Pearce-Higgins, J.W. 2016. The Breeding Bird Survey 2015. BTO Research Report 687. Thetford: British Trust for Ornithology.
Lovegrove. R., Williams, G. and Williams, I. 1994. Birds in Wales. London: Poyser.
Musgrove, A.J., Aebischer, N.J., Eaton, M.A., Hearn, R.D., Newson, S.E., Noble, D.G., Parson, M., Risely, K. and Stroud, D.A. 2013. Population estimates of birds in Great Britain and the United Kingdom. British Birds 106: pp. 64-100.
Upland bird recording and monitoring (1-day, Dalmellington, Ayr)
Brush up on your upland bird identification by songs and calls. Learn more about opportunities for participation, and practice techniques for BirdTrack and the Breeding Bird Survey (BBS). Find out about the BBS ‘Upland...