Terry Knight


In 1976 a method was developed by the Institute for Terrestrial Ecology for the recording of butterflies on a selection of about a hundred sites across the country to monitor national trends in populations. This is known as the Butterfly Monitoring Scheme. The Institute subsequently published a booklet of instructions for independent recorders so that the same method could be used elsewhere (ITE 1981). Since its publication the method has been widely adopted to record butterflies and is now used at over 400 other sites in the country.The basis of the method is as follows. A route is fixed through the site and divided into no more than fifteen sections. It is walked once a week from the beginning of April (Week 1) to the end of September (Week 26). During the walk the number of each species of butterfly seen in a section is noted. Pupae, larvae and ova are ignored just the winged stage (imagines) being recorded. Only individuals within five metres in front and about two and a half metres to each side and above ground are counted. The booklet gives more detailed instructions and limits that apply and also gives methods of analysing the data.

The site
The half mile long west facing escarpment comprising the site was purchased by the Wildlife Trust as a nature reserve in 1979 and is a Site of Special Scientific Interest. It is well suited for the monitoring of butterflies as the route and habitats are well defined. The first nine sections lie along the bridleway at the top of the bank and the remaining six are about two-thirds the way up the bank through five paddocks of calcareous grassland with some scrub (see plan). Apart from the half mile of escarpment to the south, which is in private hands, the reserve is the only such habitat of any size within a distance of about four miles. The surrounding countryside consists primarily of arable with some market gardening and intensive grazing.

Over the last thirty years, thirty two species have been seen on the reserve. The status of some of these is unclear but the best guess is as given in Table 1. Due to the difficulty of distinguishing between Small skipper and Essex skipper when in flight these are recorded as Small skipper, Small/Essex ? skipper, or Essex skipper as appropriate. The criteria used for assessing the likelihood of residence are :- larval foodplants are present in significant quantities on the reserve; imagines are seen practically every year.

Food plants
Two-thirds of the reserve is grassland comprising both fine and coarse leaved species of grass. Of these the principal ones are Tor grass Brachypodium pinnatum, fine leaved fescues Festuca rubra/ovina and Upright brome Bromopsis erecta but Cocksfoot Dactylis glomerata and Yorkshire fog Holcus lanatus , with others, are also present.

Other possible food plants include :- abundant Birds-foot trefoil Lotus corniculatus, Ivy Hedera helix, Hairy violet Viola hirta; some Bramble Rubus sp., Garlic mustard Alliaria petiolata, Nettle Urtica dioica; a little Dock Rumex sanguineus; and six Buckthorn trees Rhamnus cathartica. In recent years the amount of Bramble has declined as a result of restoring scrub invaded grassland. Garlic mustard, too, has reduced as the belt of mature scrub along the bottom of the reserve has become more open.







Nectar plants
Imagines rely on nectar and this attracts individuals onto the reserve in the summer, as well as supplying the resident butterflies. Sources on the reserve include Bramble, Violets (Viola sp.), Birds-foot trefoil, Privet Ligustrum vulgare, Wild thyme Thymus polytrichus, Pyramidal orchid Anacamptis pyramidalis, Common knapweed Centaurea nigra, Field scabious Knautia arvensis, Greater knapweed Centaurea scabiosa, Woolly thistle Cirsium eriophorum and Ivy. These flower through the season from April to November with the majority flowering from June to October. Birds-foot trefoil, Common knapweed and Field scabious are abundant.

Factors affecting numbers
The imago stage, the "butterfly" as it is usually thought of, is the one used for recording. However, this is the most mobile part of the life of the insect and individuals can travel great distances. Some such as Clouded yellow, Large white, Small white, Red admiral and Painted lady can fly here from abroad. Even species considered as resident are able to move to and from other sites. Consequently, the numbers on the reserve can be influenced by local, regional, national and even international conditions.

Factors on the reserve which will affect numbers are - availability of food plants and nectar flowers for residents, availability of nectar flowers for non-residents, a balance between sunshine and shade areas, shelter from the wind and lack of disturbance at critical times.

Normally, sheep are grazed in paddocks 1, 3, 4, and 5 (see plan) in the autumn and spring. This therefore has an effect on most of the grassland during the year and consequently on the butterflies, either directly or indirectly through the food and nectar plants. The five most numerous species (Marbled white, Meadow brown, Gatekeeper, Ringlet and Speckled wood) all feed on grasses and pass the winter ((September) October to April (May)) as caterpillars usually hidden in the base of grass tussocks during the day. As this stage in the life cycle is somewhat mobile it has the ability to cope with changes due to grazing provided this is not too severe or rapid. In order to prevent the food and nectar plants being shaded out by coarse vegetation, controlled grazing is required. This also provides areas of longer and shorter grass to suit the requirements of different species of butterfly.

Scrub clearance
Until recently, the invading scrub (mostly hawthorn) has been cleared from one paddock each year by cutting it off at ground level. As a result only about a fifth of the site has been altered per year and the resulting re-growth is at a different stage in each paddock. The effect on butterflies in the short term is likely to be very limited. However, over a longer time span just cutting off the scrub may well have had a detrimental effect as its resulting thickening and increased encroachment has reduced the area of grassland.

During the last two years more resources have been put into scrub clearance in order to halt its spread and to try to get the grassland back to its approximate 1979 boundary. This is being done by uprooting almost all the scrub back to its 1979 limit and only leaving small isolated blocks in the centre and top of the paddocks. So far nearly half of the four grazed paddocks have been done. The result, it is hoped, will be an increase in the food and nectar plants while still retaining some scrub as shelter.

At the top of the reserve is the bridleway which can be taken as one habitat (Habitat A). It consists of a strip of bare ground (mud when wet!) about one metre wide created by passing horses. Each side is bordered by short grass and herbage which increases in height away from the track. 75 metres of it has grazed paddock on one side and corn on the other, 240 metres has hedge and/or scrub both sides of it and is shaded, the remaining 485 metres has hedge and/or scrub on one side and grazed paddocks on the other giving shelter but limited shade. Some bramble and a little privet provide nectar in addition to herbs bordering the track. Over the last four years the grass and bits of the scrub along the edges of the track have been strimmed back a time or two during the summer by the County Council. Previous to this some scrub was cut back in the early 1980's but very little in the intervening years.

The second paddock from the north has not normally been grazed and consists of a mixture of long grass, discrete regenerating hawthorn bushes, permanent blocks of blackthorn scrub and some hawthorn trees. This has been taken as Habitat B. Nectar sources include taller herbs such as knapweeds, scabious, ragwort and thistles plus a small amount of bramble. The transect route through the paddock is only short being 165 metres.

The remaining four paddocks have a cumulative length of 555 metres and have been taken as Habitat C. They have normally been grazed by sheep but the southern only since 1993. As they are very over-run with Tor grass, grazing has been concentrated on the autumn, (winter), spring periods to reduce litter and to eat back the freshly emerging grass in order to allow the many herbs to flourish and flower. For practical reasons this has not always been possible and sometimes no grazing was done. Grazing has also occasionally been done in the summer thereby reducing the nectar sources available at that time.

Details of work done along the bridleway and grazing in the paddocks are given in Figure 1. The three habitat types represented can therefore be considered as in Table 2.

Butterfly densities
The maximum number of imagines recorded on one weekly count was 669 on 25th June, 2000 when 272 Gatekeepers, 150 Meadow browns, 137 Marbled whites plus others were seen. This calculates out as one butterfly seen on average every 2.3 metres (7.4 feet) of the transect with one seen on average every 1.3 metres (4.4 feet) in the centre paddock. The lowest maximum weekly count was in 1987 when only 208 were recorded.

Overall the average weekly count is calculated from 44,700 records in 26 weeks of 21 years giving 81.8 records or one butterfly every 18.6 metres (60.9 feet).

Variation by week
As might be expected, the numbers of butterflies recorded each week are much less in the colder, darker days of spring and autumn than in mid-summer. Figure 2 shows the total count of all species each week over the period from Week 1 to Week 26 for three cases. These are :- the minimum weekly count for any year, the average weekly count for all years and the maximum weekly count for any year. Each case is similar with significant numbers being recorded from week 13 (24th June) to week 20 (18th August) only.

Variation by year
Table 3 lists the fifteen most abundant species and shows their variation in numbers between 1982 (Season 1)and 2002 (Season 21). It is based on a best-fit log line and gives the back-calculated trend number in both years and the calculated compound rate of percentage increase or decrease per year. The trend numbers have been calculated by using a straight best-fit line on the graph of log(butterfly numbers) against Season and then converting the log intercepts at Season 1 and Season 21 back to actual numbers for 1982 and 2002. Calculated numbers are also given for the sum of all species and the percentage of resident species more or less confined to the grassland (Common blue, Marbled white, Meadow brown and Small heath).

Figures 3 to 5 show the detailed variation each year in numbers of the main species, all species and percentage of grassland species. Overlaid on these are the calculated best-fit log trend lines.

Variation by habitat
Over the recording period there does seem to have been a slight increase in the proportion of butterflies recorded in Habitat C (grassland) at the expense of Habitat A (bridleway). This can be seen in Figure 1 where five of the eight species appear to show this trend. It has been caused by a decline in numbers on the bridleway, an increase in the paddocks or a combination of both. The three other species (Speckled wood, Ringlet and Gatekeeper) are more associated with woodland or hedgerows and their distribution appears to have remained more stable.

National variation
National trend data are taken from The Millennium Atlas of Butterflies in Britain and Ireland (Jim Asher et al. 2001). It should be noted that these trends are from the transects recorded for the Butterfly Monitoring Scheme. As the majority of the sites monitored are nature reserves or other sites known to be important for butterflies they are comparable to Windmill Hill but not necessarily the wider countryside.

Species trends

(Figures for the variation in numbers of the most abundant species are given in Table 3 and the year by year variations are shown in Figures 3 to 5).

1 - WALL
The largest variation over the recording period is for this species which suffered a severe decline from 1983 to extinction in 1997. The cause of the decline remains a mystery but is not confined to Windmill Hill. Large areas of south-east and midland England have also been affected although since the late 1990's a slow recovery may have started.

Nationally, since 1976, this species has shown a steady decline its numbers being reduced to about a quarter by 1999. Some sites over the country, including Windmill Hill, appear to have suffered a much greater decline than others, the reason for this is not known.

The main food plants of the Large white are cultivated Brassicas (sprouts, cabbages, cauliflowers etc.). These have been, and to some extent still are, grown on an industrial scale in the surrounding Vale of Evesham. Hence, numbers on the reserve are more likely to be an indication of the quantity of local crops and effectiveness of pesticides than of conditions on the reserve itself. The significant decline probably is a local phenomenon reflecting a contraction of the industry.

There has been a significant decline on the reserve in contrast to the country as a whole where numbers have remained fairly constant. However, year by year numbers show a similar pattern of peaks and troughs as the national numbers. Hence, the reason for the decline may be due to a constant slow change on the reserve, possibly the increase in hawthorn scrub and taller summer vegetation.

The national figures indicate a drop in numbers by about a quarter which is less than that at the reserve. The reduction at Windmill Hill seems concentrated on the bridleway (Habitat A). Strimming along the edges during the summer in recent years has severely reduced the coincidence of long and short grass which is thought to be the favoured habitat of this species.

At the reserve the peaks and troughs and general increase in numbers mirror the national data. As this butterfly is likely to be a casual visitor to the reserve this is to be expected.

The year by year variation in numbers nationally and on the reserve are similar but a significant increase has occurred at Windmill Hill. This may be a result of a change in management of the grassland from sporadic burnings prior to 1979 to sheep grazing since. (It should be noted that worn specimens in flight of Meadow brown and Ringlet are difficult to distinguish and consequently some under/over recording may have occurred.)

This butterfly only occurs in small numbers at Windmill Hill and so the trend is only approximate. Figure 1 shows that the decrease is almost totally from Habitat A (the bridleway) since 1987. This would suggest that scrub invasion of the long grass along the edges is the cause. Nationally the species has increased over the recording period.

Year by year variation is similar nationally and on the reserve. However, in the former the numbers have remained fairly constant but at Windmill Hill they have increased. As this is basically a hedgerow species the increase may be a reflection of the thickening and spread of scrub on the site.

The reserve has had a significant decrease in numbers of this butterfly over the 21 years despite a small increase nationally. However, the calculated rate of decrease would be a lot less if the extremely high count of 1982 was discounted. The year by year variation is somewhat similar to the national one. As the species is probably not resident on the site, conditions in the surrounding Vale may have had an effect.

Both the year by year variation and the small decrease in numbers are similar at Windmill Hill and nationally. This is probably not surprising as it is migratory and one of the commonest species in the wider countryside.

As with the Small white there has been a very similar small decrease in numbers at the reserve. It too is one of the commonest species in the wider countryside though it is not migratory.

Since 1982 the national numbers of this butterfly have remained remarkably constant but during the 1990's its range has spread significantly. Consequently, this may hide a slight decrease in long established sites during that decade and be more consistent with the trend at the reserve. There is a considerable difference in year by year variation, with Windmill Hill showing large swings which are not reflected in the national data. This may be due in part to the difficulty of identification (see note for Meadow brown). However, as its preferred habitat is tall grass and scrub, variations in grazing and scrub clearance may have had a significant effect on its year by year numbers.

This is basically a shade loving woodland species and mostly occurs along the bridleway (Habitat A) particularly where this is bounded on both sides by hedge and/or scrub. The variation in numbers reflects the national ones although the general increase is not quite as marked. This may be because the hedge and scrub have matured over the period and become more open in the bottom, hence providing slightly less shelter. However, in the West Midlands generally the numbers do not reflect the recent national increase (Neil Gregory 2003), but show a slight decrease.

Marbled white holds the record for the highest annual transect count amounting to 995 individuals which occurred in 2000. Nationally it has had a significant increase in numbers but with a marked increase in its range. Consequently, it is difficult to relate that increase with the more modest increase at the reserve. However, the two patterns of year by year variations are similar.

These occur in small numbers and so make calculation of a trend line unreliable.

Over the 21 years the number of butterflies has remained more or less stable despite extinction of the Wall and near extinction of the Small heath due to some unknown cause affecting these two species nationally. The effect of the introduction of more concentrated management of the paddocks over the last few years appears to be beneficial as the highest numbers ever recorded occurred in the year 2000. Provided suitable management can be maintained the prospects look good for this site to remain a place to see an abundance of butterflies in mid-summer.

From 1983 to 1993 the recording of this transect was shared with Bob Woodroofe who has kindly allowed his data to be used in compiling this article. He has also given helpful comments on the draft of the article as have Neil and Corinna Gregory.

INSTITUTE OF TERRESTRIAL ECOLOGY 1981. Butterfly Monitoring Scheme - Instructions for independent recorders, NERC
ASHER,J et al 2001. The Millennium Atlas of Butterflies in Britain and Ireland, OUP
GREGORY N 2003. VC 37 (Worcestershire) Butterfly Transects Annual Review 2002, Worcestershire Record 14 : 20.


Figure 1 giving details of work done along the bridleway and grazing in the paddocks. information relating to the three habitat types are given in Table 2.

WBRC Home Worcs Record Listing by Issue Worcs Record Listing by Subject