Technically Speaking
Predation by Sparrowhawks
Effects on songbirds
by Gavin Morris
Songbird populations are affected by Sparrowhawks both by direct predation and by the effects of predation threat.
Direct predation
| Sparrowhawk population | 34,500 pairs BTO News July-August 1996 Equivalent numbers of single, unmated birds ref. BTO Atlas |
| Food consumption | Pairs plus brood anually, 2,200 sparrow-sized birds ref."The Sparrowhawk"*Singles average 2.5 sparrows-sized birds daily ref."The Sparrowhawk"* |
| Calculation | 34,500 pairs eat 2,200 sparrow-sized birds = 75,900,000 per annum 34,500 singles x 365 days x 2.5 birds = 31,481,250 sparrow-sized birds per annum |
Total sparrow-sized birds consumed by Sparrowhawks = 107,381,250 - say 107 million birds.
This method of calculation is accepted by both the RSPB and English Nature and does not include predation by winter migrants.
Predation threat - optimum fat load
Research by Oxford University Nature October 1995. "Songbirds put on fat to ensure against starvation, to fuel migration and to build up reserves for breeding. It is described as the optimum fat load. Under threat of predation by Sparrowhawks the study showed that great tits restricted their fat reserves to give better acceleration when attacked by the hawks. The same phenomenon was reported with most other songbird species which are potential sparrowhawk prey; they carried a greater fat load when the hawks were 'absent' than they do now that they are constantly under threat."
Thus songbirds are disadvantaged when feeding within hawk territories.
This finding has, however, been misused by the 'wildlife establishment' to make it appear that the Sparrowhawk is keeping the songbird population 'fit'! A typically misleading comment from a BTO spokesman who was writing in a local newspaper:
"Recent research has shown that the great tits will keep themselves in good condition without putting on fat to make themselves slow and easy pickings for the hawks."
Predation threat - effects on ranging and feeding of songbirds
Falconers are paid to fly their hawks on airfields, waste tips and in town centres, not to kill unwanted birds but to frighten them away; it is effective. Local observation since the appearance of Sparrowhawks (about 1990) is that birds no longer use their former ranges for feeding but try to find food in the safer parts of their territories near relatively safe cover. The RSPB says that this behaviour has been observed and call it 'keeping under cover' or being 'more unobtrusive' in hawk territories, but attribute no disadvantage to the prey species. Quite clearly if the ranging and feeding areas are restricted, the food available is less and the population of the prey species ultimately reduced.
When questioned on this both the BTO and the RSPB admitted, in writing, that they had no knowledge of what effect this restriction of ranging and feeding would have on the songbird populations; they continue, however, to state that:
"to maintain the impression that songbird populations are unaffected by sparrowhawk predation, other than "possible localised and particular effects". Where these localised and particlar effects might occur, or their frequency, is not stated!"
*"The Sparrowhawk" Dr Ian Newton
Questionable conclusions drawn from scientific studies
Have we been misled?
Introduction
Many hundreds of letters have been sent to national and provincial newspapers over the past decade by people who have been deeply concerned at the loss of their local songbirds through predation by sparrowhawks. Only the RSPB can know how many thousands of letters and telephone calls they have received as the sparrowhawk population has rapidly expanded across the country.
The official reply from the major wildlife organisations to such anxieties is that "Scientific Studies have shown that predation by sparrowhawks has no overall effect on songbird populations".
This paper challenges the scientific validity of these studies and concludes that predation by sparrowhawks may lead to some species becoming extinct.
The 'scientific evidence' that they quote is based mainly on three studies... at Wytham Wood, Oxford; at Brookham Common, Surrey; and from observation plots throughout Britain as analysed by the British Trust for Ornithology (with records concentrated in the London and south-east areas).
The main basis of the first two studies is the change of prey populations during 'presence', 'absence', then 'presence' of the sparrowhawk. This covers the period 'presence' when the sparrowhawk population was declining (1947-60), the period of 'absence' through a diet containing a high proportion of farmland birds with high pesticide levels (1960-72), then 'presence' again during the recovery period (1972-80) following the withdrawal of these pesticides on farmland.
Predator : Prey relationships
For sound conclusions to be drawn from studies on the effects of predation the following factual requirements must be satisfied with reasonable accuracy:
The population density of both predator and prey within the study site during the three phases of the study - an absolute requirement.
Population changes of alternative prey within, and outside, the study area during that period.
Migration rates of prey into, and out of, the study area.
Changes in predation intensity of other predators (eg weasels).
The relative vulnerability of the subject species, as compared with alternative prey species, to the predator in question (eg tree sparrow and turtle dove). Ref: 5 p.126
Changes in the supply of natural and supplied food (eg beech mast and peanuts).
Changes in natural and artifical nesting places (eg nest boxes).
Changes to the character of the study area and local environment during the period of study (eg coppicing).
Known prey preferences of the predator (in the case of the sparrowhawk prey over 20g). Ref: 7
Climatic changes (eg 1963 winter).
Predation intensity
Just as the incidence and severity of flooding depends on the duration and intensity of rainfall, so are losses of prey, through predators, dependent on the intensity and duration of predation. Records of cloudy skies and wet Sundays give no proof of past flooding, so why should irregular sightings and occasional nesting give proof of significant predation?
It is not enough for a predator to be seen once in, say ten visits by an observer to a selected area of country during a three-month period, to deduce that there has been significant predation.
Similarly, if the predator has not been seen at all during that period it would be wrong to assume it to have been 'absent' with no predation occurring. These principles particularly apply to elusive predators like the sparrowhawk.
For reliable information on predation intensity it would be necessary to census the prey within proven breeding territories of a predator and compare results with similar habitats where that predator has proved absent.
REVIEW OF STUDIES TO DATE
WYTHAM WOOD
This study was carried out by staff at the Edward Grey Institute of Field Ornithology in Wytham Wood near Oxford. It was started in 1947 to study the great tit populations in that wood. It was not specifically designed to study the effects of sparrowhawks on songbirds. The wood is about 400 ha. in extent but the study area, Marley Wood, only 26 ha. (Ref: 3 p.34)
The population of sparrowhawks was not accurately recorded but appears to have been low in the 1950s, absent until 1972, then returning to the wood after that time. These periods may conveniently be called the 'presence - absence - presence' of the hawk. (Ref: 3)
Most major wildlife organisations, and most publications state that "if predation by sparrowhawks had significant effects on songbird populations there would have been an upsurge in the tit population during the 12-year period when the hawk was 'absent'. They then claim that there was no such upsurge. Reference to fig.1 in 'The effect of Sparrowhawks on Tit Populations' (Ref: 2) clearly shows a steady increase in great tits during the decline of the sparrowhawk population up to 1960, an average increase by 50% during the absence of the hawk up to 1972, then low numbers after 1972, when the hawk returned.
How can it be said that there was no upsurge in tit numbers during the absence of the hawk?
The upsurge would have been greater if there had been no increase in predation by weasels following the loss of their main prey, rabbits, through myxomatosis, which started to take effect in 1955. To avoid losses through such predation the old wooden nest-boxes were replaced using suspended concrete nest-boxes. This avoided virtually all predation losses through, not only weasels, but also stoats, woodpeckers and squirrels. The installation of suspended nest-boxes was completed in 1976. (See also p.6 item d.)
Wytham Wood is close to suburban Oxford and local villages. From the beginning of the study to this day there has been a progressive rise in the provision of nest-boxes and nut-feeders for tits in gardens. No other species is so much favoured in this respect. To what degree has the Wytham population been supported, not only by safe nest-boxes in the study area, but also by winter food and additional nest-places outside?
It is misuse of science to apply results, however applicable to the one study species, to other ranges of species, particularly if the species being studied is atypical. The great tit is a hole-nester, is mainly single brooded and feeds mainly in shrub and tree canopy: most farmland birds have open nests and feed on the ground, in the open. Yet this study has been misused by leading wildlife organisations, and widely quoted in many bird books to prove that predation by sparrowhawks has little or no effect on songbird populations. As an example from RSPB 'Birds' magazine, Summer 1994, 'Research by the University of Oxford on Sparrowhawk Predation of Songbirds indicates only a minor effect on breeding numbers, although many may be taken.'
Scientific objection
The study subject, the great tit, is atypical of the broad range of songbird species in their ranging, feeding and nesting habits.
The study subject, as a national population, is greatly favoured by the provision of artificial food and nesting-places.
Through proximity to Oxford and local villages there is generous supply of artificial food and nesting-places close to the wood.
Safe, artificial nesting places have been provided in the study area. (Ref: 1 p.129)
The beneficial effect of preventing predation by weasels and other agents is not taken in consideration. (Ref: 1 pp.129 & 142)
The study area, Marley Wood, is too small to be representative of national populations. (Ref: 3 p.34)
Sparrowhawk population not accurately recorded (Ref: 3 p.53)
The tit population was not accurately recorded prior to 1964; this represents over half of the period of the study. (Ref: 1 p. 129)
The results were gained from an average of 30 nesting pairs of tits during the period of 'presence' to an average of 44 pairs during 'absence'. Thus a variation of only 14 pairs of an atypical species of bird has been used to demonstrate the predation effects on the whole national songbird population. See attached fig. 1 (Ref: 2 p. 133)
The study population is annually supplemented by inward migration at an average rate of 55%. (Ref: 2 p.139)
The predator population gains alternative prey from outside the study area. The effects on other species is not considered. (Ref: 2 pp. 134 & 136)
Habitat changes within the wood over 22 years not considered.
Climatic changes, for example the effects of the severe winter of 1963, not considered.
Before conclusions should be drawn from one study site, there should be replication on other sites.
Other observations
One aspect of the study showed that the breeding success of the Great Tit was greatly reduced within a 60m radius of a nesting sparrowhawk. This is equivalent to a spinney of over 1.0 ha. or 2.8 acres. Thus any wood of that size is untenable for breeding songbirds. (Ref: 5 Table 20)
Referring to 'Birds of Wytham', the following species were at satisfactory population levels, and tree-sparrows were thriving, before the return of the hawk: Tree-sparrow, redstart, tree-pipit, whitethroat, grasshopper warbler, nightingale. Some of these species, including tree-sparrows, were absent by 1990, some at low numbers.
Summary
None of the 10 minimum requirements for a study of predator/prey relationships listed on page 2, have been satisfied.
There are 14 objections to the conclusions that have been derived from this study by major wildlife organisations, every one of which, on its own, makes such conclusions unsafe.
BOOKHAM COMMON, SURREY
Introduction
Annual surveys of breeding songbird numbers, also sparrowhawk sightings, were made in this Surrey oakwood from 1949 to 1979 and recorded in the 'London Naturalist' Journals. Ref. 8. These years cover the period of 'presence-absence-presence' of the sparrowhawk, as described earlier.
The survey records were analysed by staff of the Institute of Terrestrial Ecology and their findings were published in the scientific journal of the British Trust for Ornithology 'Bird Study' vol.44. The title of the article, 'Apparent lack of impact of Sparrowhawks on songbirds' is a suitable description of the inconclusive results of this 'scientific' study.
During the period of the study, seven songbird species increased in number, two species remained stable, four declined. One species, the Starling, was absent during the early years, was common during the absence of the sparrowhawk, but was not recorded shortly after the hawk's return.
Comment
Little needs to be said to dismiss the scientific basis of this study. Whilst much of the article is devoted to the analysis of the songbird population, the status of the sparrowhawk population is dealt with in two short sentences. 'To judge from typical nest-spacing, the census plot is unlikely to have held more than one sparrowhawk pair in any one year, but Bookham Common as a whole could have held two to three pairs. This is the number known to have occurred here in the 1980s, with at least two further pairs within 1km of Bookham Common'. It should not be necessary to restate that, in a study of the relationship of predator to prey, the populations of both must be known with reasonable accuracy. Such populations should not be judged from possibilities.
If the figures for possible sparrowhawks numbers after 1974 are acceptable, then the figures from the BTO Common Bird Census for the area become important. In 1974, 63 breeding birds per ten acres were recorded, in 1979 this had dropped to 52.8 per ten acres, with a further drop to 42 in 1984. This represents an overall decrease of songbird numbers of 30% in ten years following the 'return' of the sparrowhawk. Why was this essential information not included in the study?
It would not matter if this pseudo-scientific report had been buried in some neglected library but assumptions have been widely made and incorrectly interpreted as evidence that 'sparrowhawks have no effect on songbird populations'.
Was the panel of the UK Raptor Working Group, who will be reporting to the government, influenced by incorrectly-presented material? Were they informed of the additional explanations for the apparent lack of predation effect - as set out in the final page of the 'Bookham Common Study'?
Four possible reasons for this lack of effect were listed by the authors in the following wording:-
The proportions taken from the wood were too small to be detectable in the overall counts.
Killed birds were quickly replaced from a non-territorial surplus so that the numbers of territorial (singing) birds in the wood were maintained in the face of continuing predation.
The census plot represented optimal habitat for most species, in which their densities were maintained from sub-optimal habitats where the effects of predation might have been more apparent.
When Sparrowhawks were absent, songbird mortality from other causes (including other predators) increased in compensation.
It is not mentioned in the study report but the apparent lack of predation effect could have been a combination of the four possibilities listed above.
Summary
Most telling is the comment of a senior biologist of the British Trust for Ornithology who wrote in a letter now held by SongBird Survival:
"I have read your criticism of the Bookham Common study and I agree that the study appears flawed. While there is some merit in publishing well-documented case studies, even though they represent individual instances that can be unusual, I am not sure that this is one the I would have tackled. You have raised some very damning doubts about the validity of their conclusions."
Unfortunately these conclusions were submitted to the UK Raptor Working Group as evidence.
This evidence has led to conclusions. These conclusions may result in yet further losses in British songbirds, with some species becoming extinct in some areas. The responsibility for this could lie with those who advised the panel of the UK Raptor Working Group.
REFERENCES:-
Ref: 1 Effects of predation on the numbers of Great Tits - Parus Major R.H. McCleery and C.M. Perrins (p.129) (Bird Population Studies 1991 Oxford University Press p.129-147)
Ref: 2 The effect of sparrowhawks on tit populations. C.M. Perrins and T.A. Geer. Ardea 68 (1980) 133-142 (graph p. 133)
Ref: 3 The Birds of Wytham - An Historical Survey A.G. Gosler. Fritillary 1 (1990) 29-74 (p.53)
Ref: 4 Factors affecting the delivery of prey to nesting sparrowhawks Timothy Geer. 1981 J. Zool. Lond. (1981) 195. (p.78)
Ref: 5 The Sparrowhawk. Author Ian Newton. T. & A.D. Poyser. (p.370 table 20)
Ref: 6 Apparent lack of impact of sparrowhawks on the density of some woodland songbirds Ian Newton, Lois Dale, Peter Rothery. Bird Study (1997) 44 p.129-135
Ref: 7 The complex coexistence of the Sparrowhawk and its Prey. Finnish Game and Fisheries Research Institute. 29.9.01. pekka.helle@rktl.fi
Ref: 8 The London Naturalist Journal. 1949-1980
BOOKHAM COMMON
NOTES TAKEN FROM 'THE LONDON NATURALIST' JOURNAL ON THE PRESENCE OF THE SPARROWHAWK FROM 1949 TO 1980
Dates refer to publication year giving reports of previous year's records. The notes give the whole of the information found where the Sparrowhawk is mentioned.
1949 One recorded
1950 One record in July, one in November.
1951 One record in March, two in April, one in May, one in November.
1951-9 None recorded.
The year 1959 also details - 'A Survey of Species for Bookham Common from 1954 to 1960' - The sparrowhawk is not recorded.
Under the heading 'Vagrants' may be found 'A pair of sparrowhawks nest in Central Wood and the species has been seen at the sewage farm. It is probable that these birds forage over the plains from time to time but have not been observed'.
1960-4
None recorded
1965
'A pair of sparrowhawks was found nest building on April 15th. by P.A. Alderson, and on October 13th 1963 a sparrowhawk dashed in and out of the scrub and quickly disappeared carrying a small object, which was probably a Robin but the incident was over so quickly that we could not be certain. Sparrowhawks, however, remain scarce at Bookham although there is a resident pair of kestrels'.
1966 None recorded
1967 'The Sparrowhawk has been recorded in this wood almost every Spring until 1959. Since then it has only been seen occasionally and in other seasons'.
1968 None recorded
1969 None
1970 'Sparrowhawks were seen in January, March, April, May and September'.
1971 (Report appears to be missing from bound volume)
1972 (May) This may, or may not be relevant to this examination but this entry may be found in the 'Birds' section of the Bookham Report - 'A certain amount of shooting on adjacent farmland tended to disturb the hawfinches' * A further entry 'A sparrowhawk was seen on 24th. August and one on 13th. December'.
1972 (November) None recorded.
1973 None
1974 'There were eight sightings of sparrowhawks between December 1972 and May 1973. This is a greater number of records in a short period than have been noted since regular breeding of this species ceased in 1959'. 'Regular breeding' not confirmed in annual reports.
1975 None recorded.
1976 'Sparrowhawks were resident until 1959 but were then absent until 1973' This entry clearly at odds with 1965.
There follows a table giving 'Evidence of Nesting'' recording one (nest or sighting of hawk) in the years 1949 to 1959 except for years 1952 and 1957,where there are no records. From 1960 to 1975 there are no records of nesting except for one in 1973 and one in 1975. Thus the record for 1965 has been ignored!
1976 contd. 'Regularly each Spring a pair of sparrowhawks circled in their display flight over the wood, they nested in adjoining woodland (prior to 1959). However, after 1959 they were not seen, except for an occasional bird, outside the breeding season, until 1973'. Again at odds with 1965 entry.
1977 No records.
1978 'A sparrowhawk seized a blackbird from a garden lawn close by the Common on 1st. July'.
1979 'Sparrowhawks hunted over Western and Bayfield Plains up to November, despite the finding of a female shot dead on 17th. April'.**
1980 No records
1981 No records
These records are hardly the basis for scientific evidence of the intensity of predation experienced by the prey species in Eastern Wood.
* The Bto insist that this quotation is a privately-held opinion and not neccessarily their offical view. They seem unwilling, however, to state what is their view of the validity of the study
** The evidence of shooting makes it probable that there was local control of predators including the illegal killing of sparrowhawks. This will have enhanced the survival of songbirds.