Full Terms & Conditions of access and use can be found at
ISSN: 0006-3657 (Print) 1944-6705 (Online) Journal homepage: https://www.tandfonline.com/loi/tbis20
Status of the Nightingale Luscinia megarhynchos in Britain at the end of the 20th Century with particular reference to climate change
A.M. Wilson , A.C.B. Henderson & R.J. Fuller
To cite this article: A.M. Wilson , A.C.B. Henderson & R.J. Fuller (2002) Status of the Nightingale Luscinia�megarhynchos in Britain at the end of the 20th Century with particular reference to climate change, Bird Study, 49:3, 193-204, DOI: 10.1080/00063650209461266
To link to this article: https://doi.org/10.1080/00063650209461266
Published online: 29 Mar 2010.
Submit your article to this journal
Article views: 548
View related articles
Citing articles: 16 View citing articles
The Nightingale Luscinia megarhynchos has been in decline in Britain for much of the 20th Century (Fuller et al. 1999, Parslow 1973, Holloway 1996, Huin &
Sparks 2000) and has been lost completely, as a breeding bird, in some counties over recent decades.
The two national Breeding Bird Atlases showed a marked range contraction between 1968–72 and 1988–91, with Nightingales lost from 28.5% of 10-km squares between those periods (Sharrock 1976, Gibbons et al. 1993). National Nightingale surveys in 1976 and 1980 located 3230 and 4770 singing males respectively (Hudson 1979, Davis 1982), although poor coverage in some areas would have resulted in many birds being missed in these surveys, especially in 1976.
The population was estimated at 5000–6000 pairs at the time of the 1988–91 Breeding Bird Atlas (Gibbons et al. 1993) but there has been little published infor- mation on population size or trends since this estimate.
The Nightingale is too scarce and locally distributed
to be monitored effectively by national bird monitoring schemes such as the BTO/RSPB/JNCC Breeding Bird Survey (Noble et al. 2000), the Common Birds Census (Marchant et al. 1990) or Constant Effort Site ringing (Baillie et al. 2001). County bird reports and avifaunas for counties across the Nightingale’s range have reported continuing decreases and local range contrac- tions in recent years. In contrast, a survey of Kent in 1994 located a record county total of 1066 birds and the population was estimated to be in the range 1460–1535 pairs, this higher total was probably due to better coverage than earlier surveys (Henderson 1996).
It was against this background of continued concern for the future of the species and the lack of an effective national monitoring programme to document any population changes that the BTO undertook a national Nightingale Survey in 1999.
Britain is at the northern edge of the breeding range of the Nightingale and some authors suggest that its breeding range is limited by climatic factors rather than by habitat availability (Norris 1960, Simms 1978,
Status of the Nightingale Luscinia megarhynchos in Britain at the end of the 20th Century with particular reference to climate change
ANDREW M. WILSON*, ANDREW C.B. HENDERSON and ROBERT J. FULLER British Trust for Ornithology, The Nunnery, Thetford, Norfolk, IP24 2PU, UK
CapsuleThe population level may be unchanged but the range has contracted.
AimsTo establish the current status of the Nightingale in Britain and explore causes for any changes.
MethodsOver 3000 sites where Nightingales were known to have occurred since 1980 were surveyed by volunteers between mid-April and early June 1999. A selection of 135 random tetrads were also surveyed to gauge the efficiency of the volunteer survey in locating Nightingales.
Results The survey located 4565 singing male Nightingales while the random tetrad surveys suggest that c. 32% of birds occur away from known sites, increasing the estimate for the British Nightingale pop- ulation to 6700 males (95% confidence limits 5600–9350) in 1999. A higher proportion of Nightingales was found in scrub (46.7%) than in 1976 (28.4%), suggesting a recent shift in habitat use.
Conclusion There is little evidence of a change in the size of the British Nightingale population, probably because earlier surveys underestimated numbers. The range has contracted markedly over the last few decades and numbers outside the core areas in southeast England are now low. Changes in habitat quality and increasing deer populations have caused decreases on a local scale. Changes in climate on the breeding grounds and general changes in climate or habitat suitability on the African winter quarters are likely to be important in influencing the distribution within England. Models of the effects of future climate change on Nightingale distribution in Britain predict that numbers and range should increase over the next few decades.
*Correspondence author.
A report to the British Trust for Ornithology
Ticehurst & Jourdain 1911). Its decline in Britain might therefore be linked to changes in climate on its breeding grounds but, as it is a trans-Saharan migrant, climate or habitat changes on its wintering grounds or migration routes could also be contributory factors (Fuller et al. 1999). Additionally, there have been changes in habitat on its British breeding grounds, e.g.
a reduction in traditional coppice management of woodland, exacerbated by increased grazing pressure from deer populations (Fuller et al. 1999). The abandonment of coppicing may have contributed to declines in the 19th century but reductions in Nightingale numbers during the second half of the 20th century occurred during a period when there was probably relatively little change in the amount of actively worked coppice. There has been a modest revival of coppicing in recent years, often associated with nature conservation (Peterken 1996). Further- more, it is unclear exactly how beneficial coppicing is to Nightingales because many apparently suitable coppiced woods do not support the species (Fuller 1992, Fuller et al. 1999).
Here we use data from the 1980 and 1999 Nightingale Surveys, the two Breeding Bird Atlases and county bird reports to ascertain the current breeding status of the Nightingale in Britain and investigate possible causes of recent abundance and range changes. We use climate and landscape data to assess the factors currently limiting the range of the Nightingale in Britain and climate change scenarios are then used to predict the possible effects of climate change on the species’ range.
METHODS
1999 Nightingale Survey methods
Designing a national volunteer-based survey for scarce species such as the Nightingale is problematic because random square surveys such as those adopted for more widespread species (e.g. Browne et al. 2000, Wilson et al. 2001) would be unlikely to yield sufficient data with which to produce even the most basic population estimate; and, such a survey design would result in only a small proportion of the survey volunteers finding the target species. Conversely, Nightingales are too numer- ous and widely distributed for an accurate count of all birds to be carried out annually, although this approach was attempted in the 1976 and 1980 Nightingale Surveys. It was decided that the 1999 Nightingale Survey should adopt a three-tier survey structure: (1) a
survey of all known recent sites (the Known Sites Survey); (2) a random survey of tetrads (2 × 2-km squares); (3) casual records (all records away from Known Sites).
The Known Sites Survey was, in effect, an attempt to locate all singing Nightingales. A list of known Nightingale sites was drawn up by trawling county bird reports and bird databases from county bird clubs/
societies for all sites where singing Nightingales had been heard between 1988 and 1997 (or 1998 if infor- mation was available). Additionally, all sites where Nightingales were located in 1980 were added to the list. The survey was organized through regional organizers, mainly the BTO’s Regional Represent- atives, and the fieldwork carried out by BTO members and other voluntary fieldworkers.
Volunteer observers were asked to make at least three visits to each Known Site between mid-April and early June 1999 with at least two visits during May and at least two weeks between the first and last. Observers were asked to visit the sites between nightfall (after dark) and 08:00 hours to count all singing Nightingales at each site, in at least a 1-km radius of the central grid-reference given for the site. The locations of Nightingales were marked on a sketch map and the habitat from which each bird was singing was recorded as one of ten habitat types (Table 2) adapted from Crick (1992).
The Random Tetrad Survey was designed to establish what proportion of the Nightingale population was missed by the Known Sites Survey within the main range of the species. The Random Tetrad Survey was carried-out by professional fieldworkers. This was partly to reduce the burden of survey work placed on the volunteer network and partly to ensure indepen- dence of results.
The sampling strategy was based on that adopted for the Key Squares Survey undertaken as part of the New Atlas of Breeding Birds (Gibbons et al. 1993), in which one in every nine 10-km squares on a regular grid was surveyed. With the resources available, it was estimated that nine tetrads (2 ×2-km squares) in each of 16 10-km squares could be covered. The number of Nightingales found in each of the Key Square Survey 10-km squares in the 1980 survey was calculated and the 16 with the highest number of Nightingales were chosen to maximize the chances of obtaining a reason- able sample size of tetrads holding Nightingales. Nine tetrads were selected randomly from each of these 10-km squares. The 16 10-km squares chosen in this way were representative of the current range of the
Nightingale (Fig. 1) and likely to produce a representa- tive range of results. Fieldwork methods were the same as the Known Sites Survey with three visits between late April and late May. In practice, the results of one of the 16 10-km squares were not used since volunteer survey data were not received for that square (TR03 in Kent). The failure to attain volunteer coverage in an entire 10-km square was a problem unique to TR03.
Casual records were included as part of the survey to encourage birdwatchers and members of the public to report all the Nightingales that they heard in 1999.
There was widespread publicity about the survey in the media and birdwatching press and casual records forms were widely distributed. All records were checked thoroughly to avoid duplication. The majority of birds were proven territory holders, heard on more than one occasion, although some were reported only once. A small number of records may have involved passage migrants, but obvious passage birds, such as those singing for only a day or two at coastal sites, have been excluded from the totals, as in the 1980 survey (Davis 1982).
Analytical methods
Estimating the Nightingale population size
The estimated population size of the Nightingale was calculated by comparing the counts made by volunteers in the random tetrads (combining counts at Known Sites and casual records from other areas within the random tetrads) against the total number of birds found in each of the 135 tetrads during the Random Tetrad
Survey. This gave an estimate of what proportion of the population was found by volunteers. The inverse of this was then multiplied by the number of birds found by volunteers across the whole country to give a popula- tion estimate. Ninety-five percent confidence limits for the population estimate were calculated using 999 bootstrap samples, with replacement, from the 135 tetrads surveyed (Greenwood 1991).
Estimating recent population changes
To investigate recent population changes further, counts of singing Nightingales at over 70 sites or areas where Nightingales have been counted in several successive years were extracted from annual county bird reports for the years 1981 to 1998. These included some important Nightingale concentrations (e.g.
Minsmere in Suffolk and the Church Woods complex in Kent) plus totals for a few whole counties to give a sample that was representative of the species’ range. A simple chain index was calculated (Marchant 1990) to measure population change relative to the 1981 totals.
It was decided not to include 1980 totals in this analysis as figures achieved during Nightingale surveys are likely to be higher than those reported by bird- watchers in other years.
Range limitation and climate change
The effects of climate on range limitation of the Nightingale in Britain were investigated using logistic regression in the GENMODprocedure of the SAS statis- tical package (SAS Institute 1996). The probability of Nightingale presence in each 10-km square at the time of the 1968–72 Breeding Bird Atlas was estimated using a model with binomial errors.
Presence or absence at the time of the 1968–72 Breeding Bird Atlas was used as the independent variable as this most closely followed the suggested distribution from the late 19th Century (Holloway 1996) through much of the 20th Century. This distrib- ution is therefore assumed to represent the climatically limited range in Britain before the marked range contraction subsequent to the 1968–72 Breeding Bird Atlas. The climate variables were seasonal averages for the period 1961–90 from the UKCIP98 climate change scenarios (see Appendix). Landscape variables were taken from the ITE (Institute of Terrestrial Ecology) Landcover (Fuller et al. 1994) and Landscape Characteristics (Ball et al. 1983) data sets. The former (see Appendix) providing the proportion of each 10-km square under broad habitat types: areas of scrub/orchard, deciduous woodland and coniferous
Figure 1. Distribution of 10-km squares in which random tetrad surveys were completed. ■■, Results for TR03, not used in analysis.
woodland were included as these are the main habitats occupied by Nightingales while the area of urban development was included as the distribution maps in the two Breeding Bird Atlases indicate an avoidance of large cities. An altitude variable was included from the Landscape Characteristics data set as it is well known that most Nightingales occur mainly at low altitudes in England (Simms 1978). All these data sets provided variables by 10-km square.
To generate the minimum adequate model, a step- wise selection procedure was used, variables entering the model if they were significant at P = 0.001. The null hypothesis that climate variables did not have any predictive value for the Nightingale’s range would have been accepted if none of the climate variables improved the basic model based on easting, northing and other geographical variables. The predicted climate values for each 10-km square for the period 2010–39 were substituted for the 1961–90 values in the resulting equation to obtain an estimate of the proba- bility of Nightingale presence in each square under the changed climatic conditions as predicted by the UKCIP98 climate change scenario.
RESULTS
Population size and range changes between 1980 and 1999
Over 3000 known or potential Nightingale sites were surveyed during the spring of 1999 and a total of 4498 singing males located, with an additional 67 found by surveyors of random tetrads (Table 1). Fewer Nightingales were reported in 1999 in 27 of the 32 counties in which singing Nightingales were found in the 1980 survey. Comparison of the distribution of singing males found in the 1980 and 1999 surveys (Fig. 2) shows much lower densities in the western part of the range and in the southern midlands – decreases of 70% or more were noted in Avon, Buckinghamshire, Hampshire, Hertfordshire, Oxfordshire and Wiltshire (Table 1). One of the most striking results is that the number of Nightingales reported in their ‘core range’ – the five counties of Essex, Kent, Norfolk, Suffolk and Sussex, was at least as high, and in some cases higher, in 1999 than in 1980. More than 77% of the Nightingale population was concentrated into these five counties compared with 52% in 1980 (Table 1). Numbers at
Table 1.Singing male Nightingales in each county in the 1999 survey.
Singing males Singing males Population change % of population
County in 1980 in 1999 1980 to 1999 (%) in 1999
Kent 933 1212 30 26.5
Suffolk 367 878 139 19.3
Sussex 866 692 –20 15.2
Essex 263 416 59 9.1
Norfolk 322 329 2 7.2
Berkshire 104 127 22 2.8
Cambridgeshire 199 124 –38 2.7
Somerset 181 114 –37 2.5
Hampshire 350 105 –70 2.3
Lincolnshire 159 90 –43 2.0
Surrey 98 84 –14 1.7
Dorset 120 68 –43 1.5
Gloucestershire 66 56 –15 1.2
Hereford/Worcester 66 52 –21 1.1
Bedfordshire 59 40 –32 0.9
Wiltshire 195 40 –79 0.9
Northamptonshire 45 30 –33 0.7
Isle of Wight 60 28 –53 0.6
Oxfordshire 79 17 –78 0.4
Leicestershire 8 16 100 0.4
Avon 51 11 –78 0.2
South Yorkshire 22 11 –50 0.2
Buckinghamshire 72 10 –86 0.2
Warwickshire 29 10 –66 0.2
Hertfordshire 23 5 –78 0.1
Other counties* 32 0 –100
Total 4770 4565 –4
*Devon (14 in 1980), Nottinghamshire (8), Shropshire (4), Greater London (2), Cheshire (1), West Midlands (1), Dyfed (1), Dumfries & Galloway (1).
Thorne Moors in South Yorkshire, England’s most northerly outpost appear to have been stable in recent years with 11 singing males reported in 1999 but Nightingales are now effectively extinct as regular breeding birds in other counties at the edge of the Nightingale’s range, such as Nottinghamshire, Shropshire and Devon.
The 30% increase in Nightingales reported in Kent between 1980 and 1999 is thought to be due to more effective coverage but the striking 139% increase reported for Suffolk may reflect local increases there.
The number of singing male Nightingales reported to the Suffolk County Bird Recorder for 1998 was the highest since the 1980 Nightingale Survey, with a large increase between 1997 and 1998 (Lowe 1999), although it is not known to what extent this is biased by changes in observer effort.
The random tetrad survey found 205 Nightingale territories in 135 tetrads as opposed to 138 found by volunteers at sites within these tetrads. From this we estimate that the survey located 67.3% of Nightingales, from which a population estimate of approximately 6700 territories is derived (95% confidence limits 5600–9350).
The graph of population change produced from the county bird report data suggests that Nightingale popu- lations fluctuated without any real trend for much of the 1980s but then there was a sharp decline of almost 40% between 1988 and 1991, since when populations have changed relatively little (Fig. 3).
Changes in habitat use
The habitats in which singing male Nightingales were located in the 1999 survey are shown in Table 2. These are compared with those from the 1976 Nightingale Survey (Hudson 1979) as similar data were not avail- able for the 1980 survey but data from Kent show a similar and progressive trend in changes in habitat use in surveys in 1980, 85 (East Kent only), 1994 and 1999.
These data suggest there has been a marked shift in habitat use by Nightingales in recent years. Almost half of the birds located were found in scrub in 1999 (47%) compared with just over one-quarter in 1976 (28%). In contrast, mixed woodland and active coppice are now much less important for Nightingales than it was in 1976 with the percentage of birds found in these
Figure 2.Distribution of singing male Nightingales by 10-km square (a) in 1980 and (b) in 1999.
a b
Figure 3.Nightingale population trend between 1981 and 1998.
Data from county bird reports.
habitats decreasing from 24% to 7% and 14% to 8%
respectively. There has been little overall change in the proportion of birds in deciduous woodland or carr. A chi-squared test of the changes in proportion of Nightingales in the nine habitat types in Table 2 shows a highly significant difference in habitat use between the two periods (χ2 8= 25.1, P< 0.01).
The median altitude of singing male Nightingales located in the 1999 survey was 31 metres above sea level (asl) (lower quartile 14 metres, upper quartile 55 metres). As can be seen in Fig. 4, in 1999 most Nightingales were found in areas of low altitude with 79.5% found at sites below 60 metres above sea level and none were found over 200 m asl. Unfortunately, there are no national data from previous surveys with which to make a comparison but data for Kent demon- strate a notable shift to low-lying areas. The mean
altitude of singing Nightingales in Kent in a survey in 1994 was 58.9 m asl whereas in 1999 it was 49.2 m asl, a statistically significant change (Ztest: Z= 5.97, P<
0.001). The trend is obvious in east Kent for a longer period where the mean altitude in a full survey in 1985 was 77.3 m asl but by 1999 it had reduced to 55.2 m asl;
again, a highly significant change (Z= 9.24, P< 0.001).
Range limitation and climate change
Climate variables were found to be good predictors of Nightingale range within Britain (Fig. 5a) using a binomial regression model (Fig. 5b). The probability of Nightingale presence in a 10-km square was predicted by a model including easting, northing, three landscape variables (area of deciduous woodland, area of human habitation area of land under 61 m asl) and two climatic variables (maximum summer temperature, winter potential evapotranspiration). The parameter estimates from the model are shown in Table 3. The omission of the two climate variables significantly worsens the fit of the model (χ2= 299.5, P< 0.001).
Using UK CIP98 Climate Change Scenarios to pre- dict the potential Nightingale range in the year 201039 suggests that there would be small potential range extensions northwards into Yorkshire and Cheshire and westwards into South Wales and Devon (Fig. 5b).
The model predicts that by the period 2010–39, the range of the Nightingale would have increased by 39.1% since the 1968–72 Atlas such that 889 10-km squares would be occupied. However, much of this range expansion would be due to infilling within the existing range with the probability of presence increasing markedly, rather than a large expansion into currently unoccupied areas to the north and west of the existing range.
DISCUSSION
Population and range changes
The number of singing male Nightingales reported in 1999 was just 4.5% lower than in 1980. However, no attempt was made to estimate the proportion of birds missed during the 1980 survey. It was estimated that the 1999 survey missed 32% of Nightingales and it is likely that the 1980 survey also resulted in a significant underestimate. The accumulation of a further 20 years of knowledge of Nightingale distribution locally coupled with the large increase in active birdwatchers in the years between the two surveys may have resulted
Table 2.Habitat from which Nightingales found in the 1999 and 1976 surveys were singing. Codes are those of the BTO Habitat Classification Scheme.
% of all singing males
Habitat 1999 1976
A11 Mature broadleaved woodland 21.2 22.3 A12 Coppice (active coppice only) 8.6 13.6
A2 Coniferous woodland 0.7 7.4
B5 Young plantation (less than 5 m high) 2.0
A3 Mixed woodland 6.4 23.5
A4 Carr 6.1 4.5
B1 Scrub 46.7 28.4
E Hedgerow 6.5 0
F Gardens 1.3 0.3
J Other 0.5 0
χ2 8= 26.57, P< 0.01. Totals for habitats A2 and B5 combined for chi-squared analysis.
0 5 10 15 20 25 30 35 40
<20 20–39 40–59 60–79 80–99 100–119 120–139 140–159 160–179 180–199
Territories (%)
Altitude asl (m)
Figure 4. Altitude of singing Nightingales in the 1999 Survey (n= 4392).
in fewer Nightingale sites being overlooked in 1999.
Evidence of this comes from Kent where the 1999 total was 30% higher than that for 1980, although it is unlikely that there has been a Nightingale population increase there over recent years, the higher total in 1999 being due to more thorough coverage (A.C.B.H.
unpubl. data).
The population estimate of 6700 singing males in 1999 is higher than estimates derived from either the 1980 survey or the 1988–91 Breeding Bird Atlas. Our estimate that about 32% of birds would have been missed by the survey is close to the figure of 37%
estimated to have been missed in the 1994 survey in Kent, using similar methods (Henderson 1996). Given the evidence of a decline of more than 30% since the 1980 survey (Fig. 3), these estimates, taken together with Sharrock’s (1976) estimate of 10 000 pairs in the first Breeding Bird Atlas suggest that the population may have numbered over 10 000 pairs during the 1970s.
The range of the Nightingale has undergone a marked contraction in recent decades and this appears to be continuing. In the two decades between the Breeding Bird Atlases there was a range contraction of 28.5%. Nightingales were recorded in 432 10-km squares during the 1980 survey and compared with 348 in the 1999 survey – a range contraction of 18.5%.
These figures amount to a loss of breeding range from between 1 and 1.6 10-km squares per annum. If this range contraction continues at the current rate then it is possible that population strongholds in the southeast of England will be affected within the next 20 years.
The 1999 Nightingale Survey results reveal that Nightingale numbers in their core range in the extreme southeast of England are relatively stable compared with the large decreases noted towards the edge of their British range. Range contractions towards core range areas in periods of population decrease have been noted in other species within Britain (Fuller et al. 1995) and at the continental scale in North America (Mehlman 1997). As Britain is at the northern edge of the Nightingale’s range, one might expect population decreases and/or range contraction to occur there before further south in Europe where the species is more abundant (Fuller et al. 1999). Nightingales withdrew from large parts of northwestern France since 1920 (Yeatman 1976) and became distinctly scarcer in the Low Countries and northwest Germany during the middle of the 20th century (Burton 1995), although there are as yet no signs of a reduction in Nightingale numbers elsewhere in Europe (Hagemeijer & Blair 1997). There were signs of a reversal of the long-term
Figure 5.Nightingale distribution in 1968–72 Breeding Bird Atlas (a), predicted probability of Nightingale presence in a 10-km square (b) in the period 1961–90 and (c) 2010–39.
a
c b
range reduction in northwest France (Normandy and Brittany) and The Netherlands between the mid-1970s and mid-1980s (Groupe Ornithologique Normand 1991, Hustings 1988) but the STOC-capture mist- netting programme reported a 24% decrease in the number of Nightingales caught across France between 1989 and 1998 (Julliard et al. 2001).
Range contractions of migratory species may follow a different pattern from those of residents. It is possible that sites in the southeast of Britain are occupied preferentially as they are occupied first by arriving migrants in spring. This has been suggested as a possible contributory factor in the eastward range contraction of the Turtle Dove Streptopelia turtur (Gates et al. 1994). Analysis of the arrival patterns of the Nightingale between 1893 and 1933 demonstrated that the progression of Nightingale arrivals is along a southeast to northwest axis (Huin & Sparks 2000), hence those areas occupied last are those from where the range has contracted over recent decades.
The Nightingale’s decline may be attributable to factors operating on its winter grounds, on migration and in its breeding habitats. The factors that are poten- tially responsible for its decline in Britain were reviewed by Fuller et al. (1999). The recent apparent Nightingale population decline (Fig. 3) coincides with those shown by some other species wintering in equatorial Africa. Morel & Morel (1992) list six Palearctic migrants that winter in large numbers in the rainforest zone of West Africa. These include the Nightingale and Willow Warbler Phylloscopus trochilus and two other species whose populations have declined in Britain since the mid-1980s, the Wood Warbler Phylloscopus sibilatrix and Pied Flycatcher Ficedula hypoleuca(Baillie et al. 2001, Noble et al. 2000). It is possible that common factors affecting adult survival of these species could have contributed to the decline in Nightingale numbers in Britain. Climate change, such as more frequent droughts, in these areas may be very
important but other pressures such as habitat loss and increased hunting may also play a part. Declines in adult survival rates of the Willow Warbler were probably a major cause of the observed decrease in Willow Warbler numbers in southern Britain in the early 1990s (Peach et al. 1995).
Habitat change as a possible factor affecting Nightingale abundance in Britain
There is a suggestion that habitat loss and degradation on the British breeding grounds may have caused local population decreases in many areas in recent years (Fuller et al. 1999). Nightingales in Britain depend on scrub and woodland habitats with a dense understorey;
habitat quality can be reduced through vegetation succession and development. The processes may involve reduction of actively managed coppice (although this is unlikely to be a key factor), matura- tion of post-war conifer plantations and maturation of scrub following the abandonment of traditional grazing regimes on grassland and fen early in the 20th century.
Nonetheless, the Nightingale is an opportunistic species and has been able to take advantage of newly created habitats elsewhere. Disused gravel-workings, dug mostly in post-war years, are often fringed by dense scrub, which frequently support large concentrations of Nightingales, which may have gone some way to compensating for population reductions at more traditional sites in some areas. The after-effects of the storms of October 1987 and January 1990 have certainly been beneficial to the Nightingale in Suffolk and Kent in the medium term and may have benefited the species elsewhere. Large areas of conifer plantation in the Suffolk Sandlings were devastated by the 1987 storm but the natural regeneration of scrub in some storm-damaged areas has provided ideal habitat for the Nightingale and over 60 singing male Nightingales were found in these plantations in 1999 compared with
Table 3.Climate and landscape parameter estimates from logistic regression, where the dependent variable is Nightingale presence/absence in 10-km squares in the 1968–72 Breeding Bird Atlas. Model dispersion, 0.3509; sample size, 2751.
Parameter estimate Standard error P
Easting 0.073 0.010 < 0.0001
Northing –0.101 0.010 < 0.0001
Climate effects Maximum summer temperature 1.530 0.138 < 0.0001
Winter potential evapotranspiration –10.130 1.566 < 0.0001
Landscape effects Area of land < 61 m asl 0.011 0.003 < 0.0001
Area of deciduous woodland 8.663 2.023 < 0.0001
Area of urban –6.924 1.886 0.0002
Intercept –28.442 2.845 < 0.0001
just one in 1980. The newly available habitats created by storm damage and gravel extraction will, without management targeted to retain scrub, diminish in suitability and extent in the next few years, potentially resulting in the abandonment of important Nightingale concentrations in many areas.
The recent shifts towards scrub habitats and to lower- lying areas are not easily explained but the reasons may be similar to those for the range contraction away from the north and west, i.e. an abandonment of less ideal sites as the population size diminishes. Simms (1978) suggested that Nightingales on the fringes of the species’ range in Britain are often to be found in river valleys. The 1999 Nightingale Survey results suggest that an increasingly high proportion of Nightingales are now found in very low-lying areas, especially in valley bottoms, as in Kent. Of 294 Nest Record Cards for Nightingales submitted to the British Trust for Ornithology’s Nest Records scheme between 1943 and 1980, 12% were for nests at altitudes of greater than 91 metres (Morgan 1982) but only 6.6% of Nightingales in the 1999 survey were found above this altitude.
Increased deer grazing has now become an important conservation issue in lowland woodland (Fuller & Gill 2001). Many woods within the range of the Nightingale have recently become far more heavily grazed by non-native Muntjac Muntiacus reevesiand re- introduced Roe Deer Capreolus capreolus. Muntjac have only recently become established in coastal areas of Suffolk and Sussex with very few in Kent where Roe Deer are also rare (Eric Philp pers. comm.). Intensified browsing by deer reduces the density of low woody vegetation including brambles Rubus fruticosusand tree saplings; in extreme cases it may results in a more open canopy and vigorous growth of plants close to the ground, especially grasses (Gill & Beardall 2001).
These vegetation changes generally result in a reduc- tion of habitat quality for Nightingales, which require a combination of dense low woody cover and patches of bare unvegetated ground for feeding. Examples of how deer can affect the suitability of coppiced woodland for Nightingales are reviewed by Fuller (2001) and Fuller et al. (1999). There is little doubt that deer have contributed to local population declines but it is uncer- tain to what extent they are implicated in the national range contraction of Nightingales.
Effects of climate on Nightingale distribution within Britain
Some of the variables considered in the model predict-
ing the Nightingales’ range using climate and land- scape data are highly autocorrelated and the data are also spatially autocorrelated, problems that have been demonstrated to have the potential for producing misleading results (Lennon 2000). However, the signif- icance level for inclusion of variables in this model was stringent (P < 0.001) and the model includes effects that are ecologically valid. We consider that these models do therefore support the hypothesis that the Nightingale’s range in Britain is climatically restricted, as has long been suggested (Norris 1960, Ticehurst &
Jourdain 1911) and is therefore likely to be strongly affected by future climate change, other things being equal.
The Nightingale is principally a bird of southern Europe and clearly prefers areas with a Mediterranean/
continental climate of warmer and drier springs and summers. Our model indicates that Nightingale presence/absence in 10-km squares in Britain is posi- tively associated with summer temperature and negatively associated with winter potential evapotran- spiration, i.e. areas with a more continental climate are more likely to hold Nightingales. Winter potential evapotranspiration could be crucial as this may have a strong bearing on spring soil-moisture levels. The area of deciduous woodland was positively associated with Nightingale presence while the area of human habita- tion was negatively associated, suggesting that habitat availability does affect distribution within the Nightingales’ range. The predicted future range map (Fig. 5c) shows a remarkable similarity to the range of the Nightingale in the late 19th Century and early 20th Century when the species was found as far north as mid-Yorkshire and Cheshire and was more widely distributed in southwest England and south Wales (Holloway 1996, Ticehurst & Jourdain 1911). The map of predicted Nightingale presence in 1961–90 matches the actual range in the 1968–72 Breeding Bird Atlas remarkably well (Fig. 5a & 5b). The only significant anomalies are in the north of the range where the model predicts a high probability of presence in the Lincolnshire Fens, where in fact very little suitable habitat exists, but underestimates the probability of presence in North Lincolnshire.
The Nightingale is one of few bird species with a southerly distribution in Britain to be undergoing a range contraction (Burton 1995). The northern margins of the British range of 59 southerly breeding bird species was found to have moved north by 18.5-km on average between the 1968–72 and 1988–91 Breeding Bird Atlases (Thomas & Lennon 1999). This
was attributed to the effects of climatic warming over this period. The trend of continued range contraction in the Nightingale is, therefore, at odds with the general pattern for southerly breeding species in Britain. Furthermore, the Nightingale appears not to be currently responding to climate change in the way that the spatial model presented here predicts. The Nightingale’s range has in fact contracted markedly despite the recent general pattern of global warming.
However, a closer examination shows that the recent temperature pattern has not been one of consistent increase, and temperature changes may, in fact, partly explain the variability in Nightingale numbers during the course of the 20th Century. The mean Central England Temperature (CET) for the months April to June showed a steady increase during the first few decades of the century, peaking in the mid-1940s, then decreasing steadily between the early 1960s and mid- 1980s, from when a steady increase has occurred (Fig. 6; Hulme 1999). There is much evidence from county avifaunas of a decline in Nightingale numbers in the late 19th Century, followed by a period of increase or stability for the first half of the 20th Century, with signs of an increase between 1930 and 1950, and then a more sustained decline from about 1950 (Bircham 1989, Burton 1995, Clark & Eyre 1993, Montier 1977, Payn 1978, Price 1961, Sitters 1988, Standley et al. 1996, Swain 1982, Taylor et al. 1999).
This would appear to match the periods of climate amelioration and deterioration indicated by the CET record (Fig. 6).
The apparently contradictory evidence for the effect of climate change on the Nightingale population in Britain may be in part due to the relatively short period of climate amelioration since the late 1980s.
The Nightingale is a double-brooded bird in most of its range but in Britain it is single-brooded (Snow &
Perrins 1998). It may therefore take a long time for any changes in productivity to feed though into obvious changes in population size. Further, there is evidence from the BTO’s Nest Records scheme of increase in daily nest failure rates between 1971 and 1995, although the sample sizes for Nightingales are small (Baillie et al. 2001). Even if average spring tempera- tures rise, periods of particularly inclement spring weather may increase in frequency. It is possible that breeding success may be adversely affected by relatively short periods of cold or wet weather. A further possibility is that effects of vegetation succes- sion and deer impacts are reducing the quality of habitat available in many areas with the effect that the species is unable to expand as rapidly as it might other- wise do.
Our model predicts that climatically there is the potential for the Nightingale to increase in abundance within the southern half of Britain but that predicted climate change is unlikely to make areas of northern and western Britain suitable for this species within the next few decades. An important caveat is that the model predicting the effects of climate change on the potential range of the Nightingale in Britain is rather simplistic. It assumes that land use will remain unchanged under changed climatic conditions but in reality, the responses of agriculture, leisure and development could be both considerable and complex with unknown implications for the Nightingale.
Furthermore, any widespread reduction of habitat quality through, for example deer browsing, would act to inhibit an increase in abundance and range.
ACKNOWLEDGEMENTS
We thank the many hundreds of observers who contributed to the 1999 Nightingale Survey and also the regional organisers for their efficient and effective mobilization of the volunteer fieldworkers. Chris Watts, Richard Fuller, Ian Hodgson, Andrew Plumb and Martin Sutherland carried out most of the random tetrads surveys. David Gibbons, Jack Lennon, Steve Freeman, Eric Philp, Des Vanhinsbergh and Juliet Vickery provided valuable comments and advice during the production of this paper. The Nightingale Survey was funded jointly by the Esmée Fairbairn Charitable Trust and from the BTO’s Nightingale Appeal, supported mainly by individual birdwatchers, companies and charitable trusts.
Work by professional fieldworkers was funded by the Garfield Weston Foundation. The ITE Landcover data are used with the permission of the NERC Centre for Ecology and Figure 6.Mean April to June Central England Temperature (●)
through the 20th Century (based on data in Hulme 1999). Jagged line, the five-year moving average; straight line, overall trend.
Hydrology and the Environment Agency supported the analyses associated with this paper. Finally, we thank Graham Appleton for his excellent work in raising funds through the Nightingale Appeal.
REFERENCES
Baillie, S.R., Crick, H.Q.P., Balmer, D.E., Bashford, R.I., Beav- en, L.P., Freeman, S.N., Marchant, J.H., Noble, D.G., Raven, M.J., Siriwardena, G.M., Thewlis, R. and Wern- ham, C.V.2001. Breeding Birds in the Wider Countryside: Their Conservation Status 2000. BTO Research Report No. 252. BTO, Thetford. (http://www.bto.org/birdtrends).
Ball, D.F., Radford, G.L. and Williams, W.M. 1983. A Land Characteristic Data Bank for Great Britain. ITE Occasional Paper 534. Institute of Terrestrial Ecology, Huntingdon, UK.
Bircham, P.M.M. 1989. The Birds of Cambridgeshire. Cambridge University Press, Cambridge.
Browne, S.J., Vickery, J. & Chamberlain, D.2000. Densities and population estimates of breeding Skylarks Alauda arvensisin Britain in 1997. Bird Study47: 52–65.
Burton, J.F.1995.Birds and Climate Change. A. & C. Black. London.
Clark, J.M. & Eyre, J.A. 1993. Birds of Hampshire. Hampshire Ornithological Society.
Crick, H.Q.P. 1992. A bird-habitat coding system for use in Britain and Ireland incorporating aspects of land management and human activity. Bird Study39:1–2.
Davis, P.G. 1982. Nightingales in Britain in 1980. Bird Study 29:
73–79.
Fuller, R.J.1992. Effects of coppice management on woodland breed- ing birds. In Buckley, G.P. (ed.) Ecology and Management of Coppice Woodlands: 1691–92. Chapman & Hall, London.
Fuller, R.J.2001. Responses of woodland birds to increasing numbers of deer: a review of mechanisms and evidence. Forestry74: 289–298.
Fuller, R.M., Groom, G.B. & Jones, A.R.1994. The Land Cover Map of Great Britain: an automated classification of Landsat Thematic Mapper data. Photogram. Eng. Remote Sensing 60:
553–562.
Fuller, R.J., Gregory, R.D., Gibbons, D.W., Marchant, J.H., Baillie, S.R. & Carter, N.1995. Population declines and range contractions among lowland farmland birds in Britain. Conserv. Biol.
9:1425–1441.
Fuller, R.J., Henderson, A.C.B. & Wilson, A.M. 1999. The Nightingale in England – problems and prospects. Br. Wildlife 9:
221–230
Fuller, R.J. & Gill, R.M.A.2001. Ecological impacts of increasing numbers of deer in British woodland. Forestry74: 193–199.
Gates, S., Gibbons, D.W., Lack, P.C. & Fuller, R.J. 1994.
Declining farmland bird species: modelling geographical patterns of abundance in Britain. In Edwards,P.J.,May,R.M. & Webb,N.R.
(eds) Large-Scale Ecology and Conservation Biology: 153–177.
Blackwell, Oxford.
Gibbons, D.W., Reid, J.B. & Chapman, R.A.1993.The New Atlas of Breeding Birds in Britain and Ireland: 1988–91. T. & A.D.
Poyser, London.
Gill, R.M.A. & Beardall, V. 2001. The impact of deer on wood- lands: the effects of browsing and seed dispersal on vegetation structure and composition. Forestry74: 209–218.
Greenwood, J.J.D.1991. Estimating the total number and its confi- dence limits. Bird Study39:35–37.
Group Ornithologique Normand 1991. Atlas des Oiseaux Nicheaurs Normands. GON.
Hagemeijer, W.J.M. & Blair, M.J.1997. The EBCC Atlas of Euro- pean Breeding Birds. Poyser, London.
Henderson, A.1996. Kent Nightingale Survey 1994. Kent Bird Rep.
43: 145–152.
Holloway, S.1996. The Historical Atlas of Breeding Birds in Britain and Ireland: 1875–1900. Poyser, London.
Hudson, R.1979. Nightingales in Britain in 1976. Bird Study 26:
204–212
Huin, N. & Sparks, T.H. 2000. Spring arrival patterns of the Cuckoo Cuculus canorus, Nightingale Luscinia megarhynchosand Spotted Flycatcher Muscicapa striata in Britain. Bird Study 47:
22–31.
Hulme, M. 1999. Air temperature in central England. In Cannell, M.G.R., Palutikof, J.P. & Sparks, T.H. (eds) Indicators of Climate Change in the UK. Department of the Environment, Transport and the Regions & Centre for Ecology and Hydrology, London.
Hustings, F. 1988. European Monitoring Studies of Breeding Birds.
SOVON, Beek-Ubeergen.
Julliard, R., Loïs, G., Jarry, G. & Couvet, D.2001. Oiseaux com- muns en France: variations d’abondance entre 1989 et 1998, evaluation du programme STOC-capture. Alauda 69:75–86.
Lennon, J.J.2000. Red-shifts and red herrings in geographical ecolo- gy. Ecography 23:101–113.
Lowe, G. (ed.)1999. Suffolk Birds, Vol. 48. Suffolk Naturalists’ Soci- ety, Ipswich.
Mehlman, D.W.1997. Change in avian abundance across the geo- graphic range in response to environmental change. Ecol. Applic. 7:
614–624.
Marchant, J.H., Hudson, R., Carter, S.P. & Whittington, P.
1990. Population Trends in British Breeding Birds.British Trust for Ornithology, Tring.
Morel, G.J. & Morel M.Y.1992. Habitat use by Palearctic migrant passerine birds in West Africa. Ibis 134(suppl. 1): 83–88.
Montier, D.1977. Atlas of Breeding Birds of the London Area. London Natural History Society.
Morgan, R.1982. The breeding biology of the Nightingale Luscinia megarhynchosin Britain. Bird Study29: 67–72.
Noble, D.G., Bashford, R.I. & Baillie, S.R.2000. The Breeding Bird Survey 1999.British Trust for Ornithology, Joint Nature Con- servation Committee & Royal Society for the Protection of Birds.
Norris, C.A.1960. The breeding distribution of thirty bird species in 1952. Bird Study7:129–184.
Parslow, J. 1973. Breeding Birds of Britain and Ireland. Poyser, Berkhamstead.
Payn, W.H.1978. The Birds of Suffolk. Ancient House Press, Ipswich.
Peach, W.J., Crick, H.Q.P. & Marchant, J.H.1995. The demog- raphy of the decline in the British willow warbler population. J. Appl.
Stat.22:905–922.
Peterken, G.F.1996. Natural Woodland. Ecology and Conservation in Northern Temperate Regions. Cambridge University Press, Cam- bridge.
Price, M.P.1961. Warbler fluctuations in oak woodland in the Severn Valley. Br. Birds54:100–106.
SAS Institute Inc.1996. SAS User’s Guide Version 6. SAS Institute Inc, Carey.
Sharrock, J.T.R.1976. The Atlas of Breeding Birds in Britain and Ire- land. Poyser, Berkhamstead.
Simms, E.1978. British Thrushes. Collins, London.
Sitters, H.P.1988. Tetrad Atlas of the Breeding Birds of Devon. Devon Bird Watching and Preservation Society.
Snow, D.W. & Perrins, C.M. 1998. The Birds of the Western Palearctic, concise edition. Oxford University Press, Oxford.
Standley, P., Bucknell, N.J., Swash, A. & Collins, I.D.1996.
The Birds of Berkshire. Berkshire Atlas Group.
Swain, C.M.1982. Birds of Gloucestershire. Alan Sutton Publishing.
Taylor, M., Seago, M., Allard, P. & Dorling, D.1999. The Birds of Norfolk. Pica Press, East Sussex.
Ticehurst, N.F. & Jourdain, F.C.R.1911. On the distribution of the Nightingale during the breeding season in Great Britain. Br. Birds5:
2–12.
Thomas, C.D. & Lennon, J.K. 1999. Birds extend their ranges
northwards. Nature399:213.
Wilson A.M, Vickery, J. & Browne, S.J.2001. Numbers and dis- tribution of Northern Lapwings Vanellus vanellus breeding in England and Wales in 1998. Bird Study48:2–17.
Yeatman, L. (ed.) 1976. Atlas des Oiseaux Nicheurs de France.
Société Ornithologique de France, Ministère de la Qualité de la vie Environment.
(MS received 6 October 2000; revised MS accepted 14 August 2001)
APPENDIX
Climatic variables from the UKCIP98 Climate Scenario Data Files. Values for spring, summer and winter were used in the models.
TMX Maximum daily temperature (°C) PET Potential evapotranspiration (mm/day)
PRE Precipitation (mm/day)
Landcover variables from the ITE Landcover data set. The original data gave the percentage of each 10-km square under each of 32 land cover types. The four categories shown were used for this analysis.
SCRUB Scrub/orchard
DECIDUOUS Deciduous woodland
EVERGREEN Evergreen woodland + felled forest
URBAN Urban/industrial
Landscape variables from the ITE Land Characteristics data set.
ALT1 Area of land 0–61 m asl
