Breeding status of Red-billed Choughs Pyrrhocorax pyrrhocorax in the UK and Isle of Man in 2014

ABSTRACT

Capsule: In 2014 a survey of Red-billed Choughs Pyrrhocorax pyrrhocorax in the UK and Isle of Man found substantial variation in regional trends. The population has, however, remained stable overall since 2002, with a total of 433 breeding pairs recorded.

Aims: The aim of the survey was to produce new estimates of the breeding populations and additional non-breeding birds for the Red-billed Chough in the UK and Isle of Man, and calculate national and regional population trends since previous surveys in 2002 and 1992.

Methods: A complete census was attempted, using similar field methods to previous surveys. All known breeding sites and areas of adjacent suitable habitat within the range were surveyed between April and June 2014, with additional surveys on the Isle of Man in 2015. Evidence of breeding was established based on behaviours recorded during two standard visits.

Results: The UK and Isle of Man population was estimated at 433 pairs, the majority found in Wales (55%) and Isle of Man (31%). Increases in the Isle of Man, Cornwall and south Wales contrasted with declines recorded in Scotland and in parts of north and mid-Wales. Trends differed regionally and between coastal and inland nesting areas.

Conclusions: Numbers of Choughs in the UK and Isle of Man remained stable overall between 2002 and 2014, but trends varied between geographically distinct populations, suggesting the action of diverse drivers of change. Conservation efforts need to be responsive to these pressures by maintaining and enhancing habitat and food resources where required. Where populations are small, research needs to be targeted towards identifying critical demographic stages and facilitating gene flow among adjacent populations.

Red-billed choughs Pyrrhocorax pyrrhocorax (hereafter Choughs) in Europe are at the western extent of the discontinuous world range, with populations distributed across Asia and North Africa (Cramp & Perrins 1994). In the northwest of Europe, Choughs occur in a number of small discrete populations in France (Brittany), the island of Ireland, the UK, and the Isle of Man and are recognised as being a distinct race Pyrrhocorax p. pyrrhocorax from those in southern Europe (Monaghan 1988 but see Wenzel et al. 2012). The species is found in open habitats shaped by low-intensity pastoral agriculture, predominantly in coastal areas, although inland sites are also used. Choughs are specialist feeders of soil invertebrates in these habitats and are thus susceptible to changes in land use and agricultural practices which affect abundance and accessibility of these prey.

The Chough population in the UK and Isle of Man is known to have been much more widespread prior to declines throughout the 19th century (Bullock et al. 1983). Populations on the Isle of Man and in Wales showed signs of recovery in the early 20th century but declines continued in England and Scotland. Choughs had been lost from England by 1973, by which time the population in Scotland was confined to Argyllshire (Rolfe 1966, Owen 1988, Holloway & Gibbons 2010). Historic declines have been attributed to persecution as well as land-use change, through intensification or abandonment, leading to loss of foraging habitat (Bullock et al. 1983, Owen 1988).

Three census surveys of the UK and Isle of Man Chough population have taken place: in 1982 (Bullock et al. 1983), in 1992 (Green & Williams 1992, Berrow et al. 1993, Sapsford & Moore 1994) and in 2002 (Johnstone et al. 2007). The results of each are summarised in Johnstone et al. (2007). Chough census methodology has, perhaps inevitably, evolved over time, and the differences in coverage and behaviours recorded were greatest between surveys prior to 1992. Johnstone et al. (2007) advised caution in making comparisons between population estimates over longer time periods. Comparisons of range over the longer time period are likely to still be valid due to the conspicuous nature of Chough behaviour and high levels of interest amongst birdwatchers in recording their presence. A 42% increase in the numbers was reported between 1992 and 2002 (from 202 to 429 breeding pairs; Johnstone et al. 2007). Evidence for population increase is also presented in Balmer et al. (2013) which indicated an overall increase in the range of 37% across Britain and the Isle of Man over the previous 40 years (1968/72–2007/11).

Population trends have varied regionally since the first survey and, according to ongoing annual monitoring of some populations, in the years since 2002. The population in the Isle of Man increased by 88% between 1992 and 2002. In Northern Ireland, Choughs remain a breeding species, but numbers are critically low with only one breeding pair confirmed to have bred in 2013. Small increases have been reported from the edge of the range; Glamorgan and Cornwall both held one pair in 2002 and by 2008 this had increased to six pairs (Holling 2010) and nine pairs in 2013 respectively (C. Mucklow, pers. com.).

In contrast to these increases there has been concern over declines in Scotland and in north and mid-Wales (the counties of Anglesey, Caernarfon, Denbighshire, Meirionnydd, Montgomery, Ceredigion). In Scotland, the islands of Jura and Mull have not been occupied since 1998. A single pair persisted on the mainland in 2002, but the population is now confined to Islay and Colonsay. Annual monitoring on Islay reported the loss of 14 pairs (20%) between 2002 and 2012 (Scottish Chough Study Group, unpubl. data) with the decline being attributed to a reduction in pre-breeder survival on the island (Reid et al. 2011). In north Wales annual monitoring has shown numbers nesting inland in Meirionnydd and Caernarfon have declined by 22% between 2002 and 2012 (Cross & Stratford, unpubl. data). In mid-Wales inland sites have also been abandoned over a similar period, with only one site remaining occupied in 2012 compared to ten in the same area in the early 1990s (Brenchley et al. 2013).

Research to identify drivers of regional declines indicate that the loss of the short swards favoured by Choughs that allow easy access to soil and dung invertebrate prey has occurred through both the increased grazing pressure and conversely the abandonment of grazing (Bullock et al. 1983, McCracken & Bignal 1998, Johnstone et al. 2002, Cross & Stratford 2015). An overall reduction in number of overwintered cattle has reduced the abundance of dung invertebrates (a seasonally important component of Chough diet in some areas) and use of veterinary livestock treatments have been considered as having direct adverse impacts on dung beetles and larvae (McCracken 1992, McCracken & Foster 1993). A number of hypotheses have been discussed for the overall increase in numbers of Choughs in the UK between 1992 and 2002, including a reduction in persecution as well as changes in climate, agricultural practices and Chough behaviours (Johnstone et al. 2007). Despite evidence that increasing numbers in some areas were driven by successful conservation management, including the maintenance or increasing provision of nest sites, and/or foraging habitat, the large-scale reasons for the population increase remained unclear.

The Chough remains a relatively scarce and range restricted species in the UK and Isle of Man, and while the most recent Birds of Conservation Concern assessment included the species on the Green list (Eaton et al. 2015) an assessment of extinction risk for birds in Great Britain using IUCN criteria classified the Chough as ‘Vulnerable’ due to a small, restricted breeding population (Stanbury et al. 2017). The 2014 survey was necessary to comprehensively assess any changes in breeding abundance within core areas and to resolve edge of range change in the UK and Isle of Man since 2002. We report comparisons of population estimates between 1992, 2002 and 2014 from the census surveys in the results, and go on to discuss local annual monitoring and to consider the implications of increasing survey effort and improved knowledge of breeding sites on interpretation of these trends.

Methods

The survey used a census approach, and coverage was based on known breeding locations and the historical range of the species. Specific Chough breeding locations were identified using data from a number of sources: the 2002 UK and Isle of Man Chough survey (Johnstone et al. 2007), the database of the UK Rare Breeding Birds Panel (www.rbbp.org.uk), and Bird Atlas 2007–11 (Britain and Ireland, Balmer et al. 2013). A number of discrete Chough populations are monitored annually, so liaison with local voluntary Chough Study Groups and individual volunteers was an essential part of planning for the survey to ensure no sites were missed. Information from local annual monitoring was also used as a guide to identify expected areas of stability, decline or expansion.

Surveys were targeted at 1-km squares known to have held one or more nest sites, or areas where prospecting birds had been recorded, since 1982. In areas of recent stability or decline, surveys were limited to those squares with previous records. In areas with reported increases (Cornwall and Glamorgan), surveys encompassed areas of apparently suitable habitat as well as locations with a history of occupancy.

The standard survey methodology involved two survey visits to each site, the first between 6th April and 6th May, and the second between 7th May and 15th June 2014 (as per Bullock et al. 1983). On the Isle of Man, survey coverage was incomplete in 2014; additional fieldwork was conducted in 2015 to complete the survey (Moore 2018). Observations were made at any time of day, avoiding visits during strong winds, persistent rain or poor visibility. Additional visits were made to a number of sites for the purposes of annual productivity monitoring (checking nests and ringing chicks) and data from these visits were also incorporated in the analyses presented here. The field method remained unchanged from the 2002 survey (Johnstone et al. 2007), with the exception of increasing the time watching known nests from 60 to 90 minutes, unless evidence of breeding status was obtained in a shorter period. This increase in watch time was recommended, based on information from Chough Study Groups, in order to confirm absence with more confidence. Surveyors covered the area slowly detecting calls and sightings having selected routes that optimised views of suitable nesting and foraging habitat. The linear nature of the coast essentially meant that a transect was walked taking in all suitable habitat. Surveyors also carried out watches at good vantage points. For inland sites, surveyors relied on the extensive knowledge of nest locations used in previous years and vantage point watches of suitable habitat. Visits to nests were not required for the census.

Survey work was carried out by two professional full-time fieldworkers in north and mid-Wales and one on Islay. Elsewhere survey coverage was achieved by staff from the Pembrokeshire Coast National Park Authority and the Royal Society for the Protection of Birds (RSPB), and volunteer fieldworkers associated with regional Chough Study Groups.

The count unit for the survey, breeding pairs, were classified using a well-established classification scheme of characteristic behaviours, Chough breeding status at a site can be described as confirmed, probable or possible (Appendix, Johnstone et al. 2007). The behaviour classes recorded were the same as those used in previous surveys, including observations of a pair feeding or preening each other as well as observations of single birds or a pair visiting a potential nest site. When a specific nest site was identified it was recorded as natural or artificial (quarries, mineshafts and built structures or specifically designed nest ledges and boxes) by observers in the field. Nest sites were classified as coastal or inland with inland sites defined as those within 1 km of the Mean High Water Mark (Berrow et al. 1993, Sapsford & Moore 1994). As in the 2002 survey, records of non-breeding flocks, defined as groups of three or more birds that were not family parties and were not associated with any nest sites, were also reported on the standard recording forms. In a small number of cases observers were confident to assign non-breeding status to individual birds and pairs based on behaviour and knowledge of occupied nest sites: these were also included in the non-breeding totals.

The UK and Isle of Man population estimate and regional estimates for 2014 were calculated using data including all additional information from a range of sources, including additional visits, colour ring studies and visits to nests for ringing. Additional survey visits were not all equivalent to the required two visits meaning it was not possible to correct two-visit counts using variation between standard survey visit totals and additional survey visit totals. For this reason we simply report estimates using data from the standard two visit method and present the discrepancy between the estimates.

In order to reduce the impact of varying survey intensity between years, the standard two visit results were used to calculate the population change between surveys (1992, 2002 and 2014). To estimate the numbers of non-breeding Choughs in 2014, counts of non-breeding birds during first visits (6th April to 6th May) were summed if they were either observed on the same day or were over 5 km apart when observed on different days. Johnstone et al. (2007) concluded that this approach was the most feasible and consistent method to eliminate double-counting.

Data analyses

As the survey took a census approach, population estimates were simply calculated by summing counts of possible, probable and confirmed pairs from the standard two-visit surveys and additional visits. This resulted in two population estimates for all regions and countries. The differences in population sizes between years are presented using the percentage change between survey totals from 1992, 2002 and 2014, based on the knowledge that survey methods were largely comparable between these surveys. A GLM with Poisson error distribution was used to assess variation in nest site density between countries (comparison excluded England and Northern Ireland due to the small numbers of pairs in these countries). The model was built using the glm function in the ‘sandwich’ package in Programme R and tested for significance using the anova function with likelihood ratio test. Specific comparisons between countries were calculated using Tukey’s honest significant difference tests using the glht function ‘multcomp’ package in R (Hothorn et al. 2008). All analyses were carried out in R version 3.3.1 (R Core Team).

Results

Coverage

In total, 882 previous breeding locations of Choughs were identified in 548 1-km squares (Table 1). Coverage was near complete in most areas. The methodology required two visits to be made to each site, however, some sites received only one visit, with the result that some birds may have been overlooked. Conversely, it is possible that additional birds were found on those sites which received more than the standard two visits. Seventy-five per cent (n = 661) of known sites were visited two or more times during the survey period (Table 1). Additional survey coverage varied by region, in areas where likelihood of occupation was low no additional visits were made, elsewhere around 40–50% of sites received additional surveys (e.g. on Colonsay, Anglesey and in Caernarfon). The number of visits made to 3% (n = 25) of sites was not recorded by observers.

Table 1. Distribution of Chough known historic breeding locations across UK and Isle of Man and survey coverage achieved in each region.

		Number of		Number of sites (%)
Country	County	known breeding locations	1-km squares	visited twice	(%)	visited at least twice	(%)	visited less than twice	(%)	visits unrecorded	(%)
Scotland	i. Islay	109	92	57	(52)	84	(77)	6	(6)	19	(17)
	ii. Colonsay	33	20	16	(48)	33	(100)	0	(0)
	iii. Jura	3	3	1	(33)	1	(33)	2	(67)
	iv. Dumfries & Galloway	2	2	2	(100)	2	(100)	0	(0)
	Scotland total	147	117	76	(52)	120	( 82)	8	(5)	19	(13)
Wales	A. Anglesey	76	41	19	(25)	54	(71)	22	(29)
	B. Caernarfon	168	86	25	(15)	98	(58)	70	(42)
	C. Denbighshire	5	4	0	(0)	4	(80)	1	(20)
	D. Meirionnydd	42	22	8	(19)	23	(55)	19	(45)
	E. Montgomery	9	8	2	(22)	2	(22)	7	(78)
	F. Ceredigion	54	36	8	(15)	23	(43)	31	(57)
	G. Pembrokeshire	142	109	112	(78)	138	(96)	4	(3)
	H. Glamorgan	17	14	4	(24)	6	(35)	11	(65)
	Wales total	513	320	178	(35)	348	( 68)	165	(32)
Isle of Man (2014/15)		206	99	171	(83)	180	(87)	20	(10)	6	(3)
Northern Ireland		1	1	1	(100)	1	(100)		(0)
England		13	11	10	(77)	12	(92)	1	(8)
Total		880	548	436	(49)	661	(75)	194	(22)	25	(3)

Single visits were made to 22% (n = 196) of sites, the majority of which were located in north and mid-Wales (n = 150). In many cases this was due to site occupancy being confirmed from a single visit and the remainder were not revisited as they were considered to have a low likelihood of occupancy. Sites were considered lower priority for second visits based on the history of occupancy from annual monitoring over recent years and whether the nest was known to be an alternate location within a home range of an already occupied nest. Knowledge of territory usage by colour ringed Choughs was used to determine whether there were pairs in the area that were unaccounted for.

On the Isle of Man, survey coverage was incomplete in 2014 so discrete zones (Calf, inland, north and southeast sector of south, as defined in Moore 2018) were resurveyed in 2015 and data analysed here comes from surveys which identified 120 possible, probable and confirmed breeding pairs in 2014 and a further 40 breeding pairs in 2015 (referred to as being from 2014/15 hereafter). In Cornwall/north Devon and south Wales, approximately 450 and 100 km of suitable coastline habitat were surveyed respectively.

Survey visits in the UK were carried out in between the required dates (6 April–6 May and 7 May–15 June) for 99% of first visits and 93% of second visits (equivalent data were not provided by surveyors on the Isle of Man). Data collected from visits outwith the required time periods were incorporated with that from extra visits and summarised in the results derived from using all additional information.

Survey efficiency

The number of survey visits carried out influenced the number of breeding pairs identified. Data from second visits in the UK increased the total number of pairs identified as breeding by 31% (n = 95) for confirmed and probable pairs. Where all additional information was included the number of confirmed and probable pairs identified increased overall by 7% (n = 27) (Table 2). Eleven pairs at sites classified as unoccupied based on the evidence from two standard visits were found to be breeding based on data from additional visits; evidence for confirmed or probable breeding was found for seven and possible breeding for four pairs. Data from additional visits (including colour ring records) also enable evidence of breeding to be downgraded. This was the case for ten pairs where evidence of breeding was changed from possible to absent (two on Islay and eight in Wales). The discrepancy between two visit totals and totals including all data from additional visits varies by region (Table 2).

Table 2. UK and Isle of Man population estimates and regional estimates (number of breeding pairs) from two standard visits and using all additional information. Breeding pairs were classified as possible (poss), probablr (prob) or confirmed (conf) base don the criteria given in Appendix. The discrepancy (%) between these estimates is included to show the influence of additional information on estimates (>2 standard surveys & colour ring records). The percentage of the UK and Isle of Man population in each area (based on confirmed and probable totals using all additional information) and the number of non-breeding birds are also shown.

		Number of breeding pairs based on two standard visits			Total number of breeding pairs based on two standard visits		Total number of breeding pairs based on all information		Discrepancy (%)		% UK & Isle of Man population	Number of non-breeding birds
Country	County	Poss	Prob	Conf	Conf & Prob	Conf & Prob & Poss	Conf & Prob	Conf & Prob & Poss	Conf & Prob	Conf & Prob & Poss		No.	%*
Scotland	i. Islay	8	2	44	46	54	48	52	4	−4	11	39	27
	ii. Colonsay	2	1	6	7	9	8	10	14	11	2	6	23
	iii. Jura	0	0	0	0	0	0	0			0	0
	iv. Dumfries & Galloway	0	0	0	0	0	0	0			0	0
	Scotland total	10	3	50	53	63	56	62	6	−2	13	45	27
Wales	A. Anglesey	5	2	33	35	40	40	41	14	3	9	16	16
	B. Caernarfon	14	4	74	78	92	87	90	12	−2	20	64	26
	C. Denbighshire	0	0	2	2	2	2	2	0	0	0	1	20
	D. Meirionnydd	1	1	16	17	18	18	18	6	0	4	39	52
	E. Montgomery	0	0	0	0	0	0	0			0	0
	F. Ceredigion	6	0	21	21	27	27	29	29	7	6	15	21
	G. Pembrokeshire	15	7	49	56	71	58	73	3	3	13	64	30
	H. Glamorgan	4	0	4	4	8	4	8	0	0	1	7	30
	Wales total	45	14	199	213	258	236	261	11	1	55	206	28
Isle of Man (2014/15)		23	24	108	132	155	133	160	1	3	31	111	26
Northern Ireland		0	0	1	1	1	1	1	0	0	0.2	0
England		2	1	6	7	9	7	9	0	0	2	3	14
UK and Isle of Man Total		80	42	364	406	486	433	493	7	2		365	27

*Based on counts of non-breeding birds and the sum of possible, probable and confirmed breeding pairs (multiplied by two to represent individuals).

Breeding population size and range

The census in the UK and Isle of Man in 2014 found 433 breeding pairs (493 pairs with the inclusion of records of possible breeding pairs). As shown in Table 2, this estimate, based on all available data was 7% higher than the estimate (406 pairs) using data from standard two visits alone. The largest populations were found in Wales (55%, 236 pairs) and the Isle of Man (31%, 133 pairs). The Scottish islands of Colonsay and Islay held 13% of the population (56 pairs) while southwest England (Cornwall) and Northern Ireland held the remaining 2% (7 pairs and 1 pair respectively) (Table 2).

Occupied sites were recorded in 92 10-km squares across the UK and Isle of Man (Figure 1). At a finer scale, 277 1-km squares were occupied, of which 76% contained one nest site only (based on possible, probable or confirmed breeding records). As in 2002, nest site density per 1-km square varied by country (GLM with Poisson error distribution, likelihood ratio test χ² ₂ = 23.9, P < 0.001) with the highest density being in the Isle of Man, compared to Wales (z = −4.20, df = 2, P < 0.001) and Scotland (z = −4.19, df = 2, P < 0.001).

Figure 1. (a) Geographical areas referred to in the text in relation to the current and historic distribution of Choughs in the UK and Isle of Man (Wales areas A–H: A. Anglesey, B. Caernarfon, C. Denbighshire, D. Meirionnydd, E. Montgomery, F. Ceredigion, G. Pembrokeshire, H. Glamorgan. Scotland areas i–iv: i. Islay, ii. Colonsay, iii. Jura, iv. Dumfries & Galloway, v. Mull). (b) Distribution of Chough breeding pairs in 2014. Grey 10-km squares show areas of continued occupancy in 2002 and 2014. Range loss and expansion indicated by downward and upward pointing triangles, triangles overlaid squares show net increase and decrease in number of pairs within occupied squares (light grey = 1–2pairs, mid grey = 3–5 pairs, dark grey = 6–8 pairs).

Nest sites occupied

In Scotland, Wales and Isle of Man 483 nest sites were located in 2014, and these were predominantly in coastal areas (86%, n = 414, Table 3). In England, all seven nest sites were coastal, due to the small sample size they were not included in subsequent analyses. The remaining 14% (n = 69) of nest sites were classified as inland (within 1 km of the coast). The use of coastal and inland sites varied between countries (χ² ₁ = 13.24, df = 2, P = 0.001); Scotland had the largest proportion of Choughs nesting at inland sites (27%), compared to 14% and 9% of the Welsh and Isle of Man populations respectively. No inland sites were reported in England or Northern Ireland.

Table 3. Nest location and nest type in Scotland, Wales and Isle of Man occupied by confirmed, probable and possible breeding pairs of Choughs and % change between 2002 and 2014 by country.

Nest location	Country	Scotland (%)	Wales (%)	Isle of Man (%)	Total (%)
Coast	2002*	68 (82)	205 (82)	142 (95)	415 (86)
	2014	45 (73)	223 (85)	146 (91)	414 (86)
	% change	−34	9	3	0.5
Inland	2002*	15 (18)	46 (18)	8 (5)	69 (14)
	2014	17 (27)	38 (15)	14 (9)	69 (14)
	% change	13	−17	75	0
Nest type
Artificial	2002	33 (40)	67 (27)	18 (12)	118 (25)
	2014	41 (67)	71 (27)	31 (19)	143 (30)
	% change	24	6	67	20
Natural	2002	50 (60)	178 (73)	132 (88)	360 (75)
	2014	20 (33)	190 (73)	129 (81)	339 (70)
	% change	−60	8	−2	−5

*2002 figures for nest location vary slightly from Johnstone et al. (2007) due to improved knowledge of specific nest sites and differences in GIS methods used to identify coastal and inland sites.

Of all nest sites identified in 2014, 29% (n = 143) were reported as artificial (built structures or nest ledges/boxes) (Table 3). Use of artificial nest sites varied by country (χ² ₁= 51.43, df = 2, P < 0.001) and by coastal/inland location within each country (Cochran-Mantel-Haenszel χ² ₁ = 123.8, df = 1, P < 0.001). The clear pattern is that the majority of inland nest sites (90%, n = 61) were artificial compared to only 19% (n = 81) of coastal sites, in the absence of full figures on the availability of artificial nest sites on the coast it is not possible to determine whether this shows a preference or simply reflects the greater abundance of artificial sites inland.

Comparison with population size from previous surveys

In order to reduce the effect of survey activity varying between 1992, 2002 and 2014, population change was calculated using confirmed and probable totals using the standard two survey visit method. This indicates that the Chough population in the UK and Isle of Man, estimated at 406 pairs using just two visit data, has increased by 34% since 1992 (302 pairs) and by 3% since 2002 (395 pairs) (Table 4).

Table 4. Change in numbers of Choughs breeding in UK and Isle of Man between surveys 1992, 2002 and 2014. Percentage population shown in brackets. Grey shading indicates counties referred to in text as north and mid-Wales.

		Confirmed and probable breeding pairs based on two standard visits						% change
Country	County	1992^1 (%)		2002 (%)		2014 (%)		1992–2014	2002—14
Scotland	i. Islay	69	(23)	56	(14)	46	(11)	−34	−18
	ii. Colonsay	9	(3)	14	(4)	7	(2)	−22	−50
	iii. Jura	2	(0.6)	0	(0)	0	(0)	−100
	iv. Dumfries & Galloway	1	(0.3)	1	(0.3)	0	(0)	−100	−100
	v. Mull	1	(0.3)	0	(0)	0	(0)	−100
	Scotland total	82	(27)	71	(18)	53	(13)	−35	−25
	A. Anglesey	13	(4)	37	(9)	35	(9)	169	−5
	B. Caernarfon	56	(18)	80	(20)	78	(19)	41	−3
	C. Denbighshire	0	(0)	2	(0.5)	2	(0)	Increase	0
	D. Meirionnydd	8	(3)	16	(4)	17	(4)	113	6.3
	E. Montgomery	4	(1)	1	(0.3)	0	(0)	−100	−100
	F. Ceredigion	15	(5)	25	(6)	21	(5)	29	−16
	G. Pembrokeshire	53	(17)	46	(12)	56	(13)	6	22
	H. Glamorgan	1	(0.3)	1	(0.3)	4	(1)	300	300
	Wales total	150	(50)	208^2	(53)	213	(55)	42	2
Isle of Man		68	(22)	114	(29)	132	(32)	94	16
Northern Ireland		2	(0.6)	1	(0.3)	1	(0.2)	−50	0
England		0	(0)	1	(0.3)	7	(2)		600
UK and Isle of Man Total		302		395^2		406		34	3

¹1992 estimates: regional totals for Scotland – unpublished data and Wales – Green & Williams 1992. Isle of Man, Northern Ireland and England as published in Johnstone et al. (2007).

²Two visit 2002 totals for UK and Isle of Man and Wales are four fewer than that reported in Johnstone et al. (2007) due to duplicate nest sites being erroneously included.

Population change varied across the UK and Isle of Man between the three surveys. There has been no substantial change in Wales since the most recent survey (3%), although trends have varied regionally. Between 2002 and 2014 numbers declined Ceredigion, Anglesey and Caernarfon and no Choughs were found to be breeding in Montgomery in 2014. In contrast, numbers in Pembrokeshire were up and in Glamorgan 4 confirmed pairs were reported in 2014, compared to a single probable pair in 2002. The Chough population on the Isle of Man has increased by 94% since 1992, when it was understood to be 68 pairs. Since 2002 the number of confirmed and probable pairs recorded (2014/15) was 16% higher. The population in England has increased from a single breeding pair present in 2002 to seven breeding pairs in 2014.

The population in Scotland declined by 25% between 2002 and 2014. The entire population is now restricted to the islands of Islay and Colonsay. On Islay the population has declined by 18% since 2002. On Colonsay numbers have been more variable, increasing from 9 to 14 pairs between 1992 and 2002; in 2014, 7 pairs were reported and 6 of 10 previously occupied sites in the northern part of the island were found to be vacant.

Across the UK and Isle of Man the number of pairs of Choughs on the coast and inland has shown minimal change since 2002; a 0.5% increase on the coast and a no change inland (Table 3). However, trends do vary by country. In Wales, numbers of Choughs on the coast have increased by 10% compared to a 17% decline inland. In contrast, numbers on inland sites in Scotland have increased by 13%, but on the coast they have declined by 34%. Nest type varied overall between 2002 and 2014 (Cochran-Mantel-Haenszel χ² ₁ = 4.6, df = 1, P < 0.032), driven largely by the proportion of artificial nest sites used increasing in Scotland from 40% to 67% (Fisher test P = 0.001).

Numbers of non-breeding choughs

A total of 365 non-breeding Choughs were recorded during the first visit period in 2014. Non-breeding birds were present in all populations except Northern Ireland, representing between 14% and 52% of birds in each (Table 2). Combining these counts with the 493 breeding pairs gives a total of 1357 individual Choughs in the UK and Isle of Man, of which 27% were non-breeding birds. This estimate indicates that the overall proportion of non-breeding birds in the UK and Isle of Man population does not appear to have changed since 2002 when, using the same approach the proportion was also 27%. There was variation between some populations; the proportion of non-breeding birds dropped from 35% to 26% in Scotland, notably on Colonsay where non-breeders made up 45% of the population in 2002 compared to 23% in 2014. In Wales, the proportion of all birds that were non-breeders recorded increased slightly from 21% in 2002 to 28% in 2014.

Discussion

We found 433 (confirmed and probable) breeding pairs of Choughs in the UK and Isle of Man in 2014. The majority of the population was in Wales (55%) and the Isle of Man (31%). Comparing results based on two standard survey visits, the population showed little change (3%) since 2002 and was 34% higher than 1992, although substantial variation in regional population trends was apparent. Chough populations across the range of the northwest European race have not been surveyed comprehensively since 2002 (Gray et al. 2003, Johnstone et al. 2007). In 2014 surveys were carried out in Brittany (excluding Belle-Ile) using the same methods as reported here. Results indicate an increase in numbers in this area; 39 pairs in 2014 compared to 30 pairs in 2002 (Laurent Gager, unpubl. data). Combining UK and Isle of Man census data with data from Brittany (excluding Belle-Ile) gives a minimum population estimate of 474 territorial pairs in northwest Europe; 4% higher than in 2002.

The largest increases in population since 2002 were in Pembrokeshire in south Wales and on the Isle of Man (comparing results from two standard survey visits). In Pembrokeshire, however, the longer term pattern shows numbers have been stable since 1992. The increase in the Isle of Man was in part due to the recolonisation of the Northern Plain but there have also been moderate increases in all other parts of the island except the ‘Western Zone’ and on the Calf (Moore 2018). Two regions, Glamorgan in south Wales and Cornwall in England, indicate the potential for further colonisation of suitable habitat by this species.

Survey accuracy

The field methods and the subsequent interpretation of behaviours used in the 2014 census followed recommendations from the previous survey. Johnstone et al. (2007) assessed the sensitivity of the number of survey visits and behaviours used to assign confidence in breeding status and concluded that using all additional information over and above the standard two visits had a limited effect on the total number of breeding pairs identified. Likewise, in 2014 such additional information increased by 7% the number of confirmed and probable breeding pairs identified. Confirmation of breeding status which may have been wrongly classified by two visits, was the clear benefit arising from additional visits. This was the case in parts of north and mid-Wales where the discrepancy between standard visit and totals based on all additional information in one region was as high as 29%. Nevertheless, while population estimates can be improved by additional visits, the standard two survey visits should be considered adequate for the purposes of a national survey in order to estimate population size and as a standardised method to calculate trends.

Incomplete coverage on the Isle of Man in 2014 mean that data from 2015 have been incorporated in these results. There was the potential that revisiting areas in the subsequent year could inflate the counts by surveyors actively seeking to find pairs that they ‘missed’ in the previous year. To minimise this risk, instead of targeting individual nest sites, discrete zones were resurveyed in 2015.

Context provided by ongoing annual monitoring

Context can be provided to single year census surveys by comparing results with those from regional/local annual monitoring. Annual monitoring and colour ringing in Wales and Scotland has substantially improved our understanding of Chough population dynamics and distribution over the last three decades (Bignal et al. 1997, Cross & Stratford 2015).

In north and mid-Wales, despite increasing coverage each year (since c.1994) numbers in the core of the range show a drop of ten pairs since 2002, although they remain higher than in 1992. Annual monitoring has considerably improved knowledge of Chough nesting locations across this large area. Twenty-two nest sites that were first located in the late 2000s are considered by surveyors to have been occupied for some years prior to discovery, based on the presence of a build-up of old nest material. The implication of this is that results from previous surveys may have underestimated the population in this region. Assuming occupancy of these 22 sites was similar to the proportion of occupied sites across the region, up to 14 additional pairs may have been undiscovered in 2002 (Cross & Stratford, unpubl. data). Adding these to the published figure for 2002, and comparing against confirmed, probable and possible records from 2014 gives a decline in north and mid-Wales of 11%, compared to the decline of 5% reported in this paper. The implication of this would be that the trend for the whole of Wales would be slightly negative (−4%) and the UK and Isle of Man trend would be −2%.

The number of Choughs in Scotland has declined consistently since 1992, with ten pairs lost on Islay and seven on Colonsay since 2002. This decline has led to a number of research projects to identify causes and potential solutions (Bignal et al. 1997, Reid et al. 2009, 2011). Annual monitoring from Colonsay shows more fluctuation compared to the consistent decline on Islay. The population on Colonsay was stable at around ten pairs during the 1990s having increased steadily since the early 1980s following recolonisation in the late 1960s (Finney & Jardine 2003). Numbers may now be stabilising around this level again (9 pairs, including possible records, in 2014 and 2015; Holling 2017) following a period of growth peaking at a maximum of 18–19 pairs around 2005–08, and subsequently declining (D. Jardine, unpubl. data).

Future monitoring of Choughs in the UK and Isle of Man would benefit from annual surveys of a representative sample of 1-km squares to improve our ability to detect change in population size over time and provide the data required for analyses to determine national and regional drivers of change. The sampling protocol would need to ensure a proportionate level of coverage is maintained across regions, and include squares with known occupation as well as areas of suitable habitat with no history of breeding in order to detect population expansion. In addition, an approach to ensure detection of increasing nest density within the existing range would be required.

Drivers of change

The varied trends in different parts of the UK and Isle of Man Chough population indicate that despite numbers appearing stable overall, different drivers may be operating with differing intensity across the range with population responses varying spatially and temporally (Reid et al. 2008). While it was beyond the scope of this paper to define the specific drivers of population change across the UK and Isle of Man, we highlight here the evidence for the likely factors involved which will hopefully be useful in directing future work.

The distribution of Choughs in the UK and Isle of Man has been described as being limited to areas with mild winters, suitable nest sites and grazed pastures (Monaghan 1988). Breeding success, survival and breeding longevity of Choughs has been shown to be correlated with weather conditions the preceding year via the impact on invertebrate abundance, productivity and availability (Reid et al. 2003, 2008). The northwest European population, being at the northern edge of the global range, is projected to have the potential to benefit from future climate change bringing warmer temperatures (Huntley et al. 2007). Given current evidence of climate change in the UK to date (Humphrey & Murphy 2017) it is possible that conditions in southern England and south Wales (and in Brittany) may have been more beneficial for Choughs in recent years. This may have contributed to the population increases reported here through enhanced breeding success and survival.

Spatial variation in future climate change may result in variation in its impact between discrete Chough populations. This is especially true for rainfall patterns which, in Scotland in particular, are projected to include a larger proportion of annual rainfall falling in the form of heavy rainfall events (Humphrey & Murphy 2017), which could have a negative impact on Choughs by reducing invertebrate activity and abundance (Reid et al. 2003, 2011).

The dependence of Choughs on low-intensity pastoral farming systems makes populations vulnerable to agricultural change, particularly in relation to grazing, which can alter the availability of short grassland vegetation and availability of dung invertebrates (Bullock et al. 1983, Berrow et al. 1993, Bignal et al. 1997, Johnstone et al. 2002). Regional Chough declines during the late 20th century have been attributed to reductions in grazing of cliff-top foraging areas and loss of permanent coastal pastures, as well as cessation of overwintering cattle outdoors (Bullock et al. 1983).

In Wales, Chough numbers inland declined (−17%) while numbers at coastal sites increased (+10%) between 2002 and 2014. Inland declines in north and mid-Wales, where the majority of inland sites are located, appear to be associated with loss of short sward pastures near nest sites through reductions in upland grazing (14% decline in sheep numbers in Snowdonia between 2002 and 2010; Silcock et al. 2012). The only options for management of unenclosed habitat under the current agri-environment scheme in Wales are associated with heather moorland restoration and therefore only limited grazing is allowed (Welsh Government 2017).

Likewise, fluctuations in numbers of Choughs on Colonsay are also suspected to be linked to levels of grazing on the island. Grazing increased during the mid-2000s when the population peaked and subsequently declined again to similar levels present in the 1980s and 1990s; as a result, some important grassland areas have now become less suitable for Choughs (D. Jardine, pers. comm.). Specific changes in management of machair for botanical objectives in the north of the island may have impacted on the mosaic of habitats available to Choughs in this area (Calladine et al. 2014).

Maintaining and restoring foraging habitat, particularly in areas where local population size is considered too small, has been the focus of much conservation action. At a local scale, conditions around nest sites and connectivity between breeding sites and surrounding foraging areas have a long-term impact on fledging success and lifetime reproductive success (Reid et al. 2003, 2006). The impact of habitat provision has been quantified in Brittany (Ouessant) where Choughs fledged one extra chick for every additional hectare of short sward (<5 cm) within 300 m of a nest (Kerbiriou et al. 2006). Choughs are known to exploit seasonally available prey that may fluctuate annually (cereals grains, bilberries, ants and beetles) so maintaining a diverse habitat matrix is important to ensure varied foraging opportunities through the year (Bignal et al. 1997, Johnstone et al. 2002, Reid et al. 2009).

Even where suitable grazing levels are maintained, experiments have shown significant reductions in the invertebrate larvae of species important to Choughs in the dung of cattle treated with common animal health treatments (Gilbert et al. 2018). The invertebrates affected, namely beetle and fly larvae, are predominant food sources for Choughs nesting on the coast and in late summer on Islay (Bignal et al. 1997). Declines on Islay were greatest on the coast compared to inland sites, matching modelled estimates of population growth rate for these regions (Reid et al. 2006). There is a real possibility that that the veterinary treatment of livestock may have contributed to the reduction of food supply for Choughs (Gilbert et al. 2018) given the current low level of pre-breeding survival on Islay (Reid et al. 2004, 2011, Bignal & Bignal 2011) and differences between food preferences and population trajectories on the coast compared to inland sites. While the research was carried out on Islay the findings are likely to be important wherever Choughs are reliant on stock grazed habitat. These veterinary treatments have been recommended not to be used on farms where Choughs feed (European Commission 2016). However, demand for such treatments to be used is predicted to increase with increased treatment resistance and with climate change leading to increasing incidence of parasitic problems in livestock (Van Dijk et al. 2010, Gordon et al. 2012).

For a long-lived species with a complex social structure, changes in seasonally important habitat availability can have a disproportionate impact. For pre-breeding birds, foraging flocks and communal roosts act as information centres, facilitating discovery and tracking of shifting food supplies as well as mate selection (Bignal et al. 1997, Still et al. 1987). In areas of population expansion, reports of the development of seasonal activities of pre-breeding flocks have been viewed as a positive sign indicative of the emergence of a social structure similar to that seen in the core of the range (Johnstone et al. 2011). Conversely, factors that reduce availability or quality of important flock foraging locations could exacerbate declines by interfering with these social processes (Bignal et al. 1997, Blanco & Tella 1999, Wright et al. 2003).

Concerns have also been raised about population viability in terms of genetic diversity particularly in the northernmost populations. Studies of Chough samples from across Europe described a north–south gradient in genetic diversity with Scotland and Isle of Man having the lowest diversity (Wenzel et al. 2012). Long-term inbreeding within the Islay population has led to lethal blindness in nestlings caused by a large-effect lethal recessive allele providing evidence that the effective population size is critically small (Trask et al. 2016, 2017). Currently low levels of gene flow do occur naturally. For example, dispersal and subsequent breeding has been recorded between north Wales and the Isle of Man (Moore 2006, 2008). Genetic management as a conservation intervention on its own would of course be insufficient to conserve the Scottish Chough population; much has been written about the habitat management required to maintain, improve and increase availability of foraging habitat and prey populations for Choughs (Johnstone et al. 2002, Reid et al. 2008, 2009).

Conclusion

National species surveys using standardised methodology provide important snap shots of status across time. Considering results from national surveys alongside more intensive regional annual monitoring help to validate and provide context to the findings of these periodic large-scale surveys. While the UK and Isle of Man Chough population has remained stable since 2002, trends have varied regionally. Broad environmental changes (e.g. climate and agriculture) interact with regionally distinct changes north to south across the range. There is also evidence that the northern most and genetically least diverse population may be particularly vulnerable due to the effects of inbreeding.

Choughs occur in landscapes with high cultural value, which bring opportunities for their conservation. Managing and valuing the pastoral landscape and traditional farming practices may provide opportunities for Choughs. The perceived conflict between food production requirements and the needs of other species groups can be dealt with through careful appropriate grazing management to maintain areas of low-intensity pasture for botanical and invertebrate interests in such a way that is very much compatible with the needs of species such as Chough.

Despite the diverse land-use practices across their UK range there is commonality across the issues and Choughs provide a particular opportunity relying as they do on the habitats now maintained by low-intensity pastoral farming. There is potential to identify specific causes of decline by comparing candidate drivers of change between populations with different population trajectories, thus enabling the targeting of particular solutions and conservation approaches most effectively.

Acknowledgements

Particular thanks go to the Scottish Chough Study Group, the Manx Chough Project, Manx BirdLife, the Manx Ornithological Society, the Manx Ringing Group and the Calf of Man Bird Observatory, Pembrokeshire Coast National Park Authority staff and local volunteers and coordinators Bob Haycock and Jane Hodges, to Tony Cross and the Cross and Stratford Welsh Chough Project, to Claire Mucklow and Cornish Chough volunteers, to Mike Peacock and David Jardine and to RSPB reserve staff who all contributed data to the survey. Also to Laurent Gager, Christine Blaize, Christian Kerbiriou and volunteers for Chough counts from Brittany. We are very grateful to the landowners on whose land the survey was conducted. Reviewer and editorial comments improved an earlier draft of this manuscript.

Additional information

Funding

We would like to thank NRW and RSPB for funding this survey in the UK as part of the Statutory Conservation Agencies & RSPB Annual Breeding Bird Surveys (SCARABBS), and to the Manx Chough Project for funding the survey in the Isle of Man.

Appendix. Evidence used to classify breeding status for Choughs

Table

Behaviour	Code	Breeding evidence
Single on ground	SG	NONE*
Single flying	SF
Pair flying	PF
Pair on ground	PG	POSSIBLE
Single birds; one in, one out	1-1
Pairs; two in, two out.	2-2
Carrying nest material	CM	PROBABLE
Male feeding female	MF
Mating	MT
Visits to potential nest sites One in, two out	1-2
Nest with eggs	NE	CONFIRMED
Visits to potential nest sites Two in, one out	2-1
Visits to potential nest sites One in & stays in	1-0
Bird carrying faecal sac or eggshell	FE
Nestlings heard begging	NH
Nestlings seen in the nest	NY
Fledged young	FY
Family group	FG

These characteristic behaviours were recorded during census visits and used to systematically assign sites to categories, based on the level of evidence for breeding. Single birds flying (SF) or on the ground (SG) and paired birds flying (PF) were recorded but were not sufficient evidence of breeding activity. Paired birds often feed in unison and allopreen. A pair on the ground (PG) is weak evidence of a breeding attempt because birds may pair in the years before they first attempt to breed. Visits to potential nest sites may be made by single birds (1-1) or pairs (2-2) and provide stronger evidence of site occupancy. A bird carrying nest material (CM) (usually twigs or sheep’s wool) is good evidence that a breeding attempt is underway. This also holds for mating (MT) and courtship feeding (MF). Females beg by calling while crouched with quivering wings (care was taken not to confuse this with adults feeding young, which do not have bright red legs and bill). Some nests, such as those in buildings, were accessible and could be checked under licence for clutches of eggs (NE). Choughs can visit nest sites in different ways. One in, two out (1-2) represents the male returning to a nest and enticing the incubating female off to feed her. Two in, one out (2-1) represents the pair entering a nest site together, the male leaving once the female has settled on eggs. More usually, the female returns alone to resume incubation (1-0). These behaviours are strong evidence that breeding is underway. A bird carrying a faecal sac or eggshell in its bill (FE) is good evidence that young are present. Nestlings may also be heard begging (NH) or seen in the nest (NY). For a week after fledging, young Choughs call while hidden near the nest, emerging when their parents return to feed them (FY). Care was taken not to confuse these temporary refuges with nests missed on earlier visits. After fledging, feeding adults are often accompanied by begging young in a family group (FG). These young have darker legs and bills for several weeks after fledging; although the colour of the bill and legs brightens within the first month, they can still be distinguished in flight by shorter, rounder wingtips. These behaviours were classified into levels of evidence for breeding having taken place.