A look at Mortality Rates of Sparrow Hawks in Scotland


An investigation in South-East Scotland, spanning eleven years, studied predation by Eurasian sparrow hawks on density-dependent mortality of wintering redshanks. Research was undertaken by Philip Whitfield of Scottish National Heritage, along a 5km rocky shore at Scoughall, on the pouter Firth of Forth. The researcher studied wintering redshanks for winters 1983 to 1993 [August- April].

Considerations were given for the impact of the predation on prey populations, which are controlled by changes in prey density. Whitfield was interested in whether predation compensates for other influences on prey abundance or whether it has additive effects.

Older, previous studies had shown those predators taking prey had no, or little, effect on populations. [Goransson 1975] However, more recent studies showed a variable effect of predators on prey populations. [Kenward 1977, 1986, Kenward, Marcstrom & Karlbom 1981] Latter studies concluded that density dependent predation regulated prey abundance and produced stability in prey populations.

In shorebirds attention was focused on density independent effects, especially influences of severe weather on winter survival. This was one of few studies done on density-dependent processes in shorebirds. Reasons for this are the high mobility, which causes difficulty in measuring mortality. Density-dependent processes are also difficult to isolate and population size may not be variable enough over a given period to detect density-dependence.

On some sites in Britain, redshank mortality is high due to severe weather, whereas other sites are high-risk predatory areas. One such site is the investigative area at Scoughall. Here density independent mortality from severe weather is rare. Redshank population is stable within each winter and death from sparrow hawk predation can be measured through recovery of kills. Observations on the site were conducted throughout the winter, where records were taken at high tide when birds were arriving or leaving roosts in restricted feeding areas. It was assumed that the research site remained constant over the period of study and a maximum of redshank was undertaken annually, in early October, and used as the relevant population density. This was a stable population since most birds stayed until April migration. Birds were considered juvenile where it was their first resident winter and adult in their second year.

Searches of the shoreline and adjacent inland areas were undertaken for the recovery of dead redshanks and classified as sparrow hawk kills or other. Remains of sparrow hawk kills were found to be distinctive and easily separated from kills by other predators. [Cresswell 1995] Whitfield’s conclusion was based on discovery that 95% of remains were left by sparrow hawks. Not all deaths were observed and in those that were not, an estimation of the date of death was based on previous searches of the area together with the appearance of the remains. This process was found to be accurate within one week of the death date.

The study found that most sparrow hawks initially carried their prey to the closest shoreline cover. However, the dead redshank could then be taken to a number of different locations. This caused a risk in double counting of corpses and so data was only taken where body parts were common. For example, where it was suspected there might be more than one corpse, bodies were only counted as more than one if two right legs were evident. Juvenile and adult birds were analysed separately because of age dependent mortality. Density and mortality were measured using different processes.

Weather data was taken from the British Atmospheric Data Centre. This data was based on daily surface records 5km south of the study site. Data were used to study the effect of winter weather on redshank mortality. Different variables were chosen because of previous studies, which showed increased risk of predation from sparrow hawks in wind and rain [Hilton, Ruxton & Cresswell 1999] It was expected that mortality would be affected by rainfall but not by wind speed.

Highest redshank mortality occurred between October and February. Although this period of the year generally suffers the worst of the weather, few strong relationships were found between the weather conditions and monthly mortality. No association to weather and mortality was found in the month where highest mortality occurred. From December to February mortality rates in both juvenile and adult classes were especially high in 1989 although Whitfield did not consider this influential in regards to the weather.

Density-dependent winter mortality was found in both juvenile and adult redshanks due to predation by sparrow hawks. Competition for food on the sea side of the strandline forced individuals to the strandline, where predation risks are higher. The strandline, being closer to the scrub, ensures cover for sparrow hawks who await ready to take prey. The higher the number of redshank on the strandline, the higher the mortality rates from predation. Territorial behaviour from adult redshanks pushes juveniles to the strandline and into higher risk predation areas. Furthermore, higher numbers of redshank at the strandline attract higher predation from sparowhawks because raptors respond to prey density [Holling 1959, Sonerud 1992]

Whitfields study showed that the major cause of density-dependant mortality of prey came from a density-mediated change in prey behaviour, where redshank made themselves vulnerable in high risk/high feed areas, even though the areas were not preferred by redshank. Previous studies showed that prey numbers were not regulated by sparrow hawk predation. [Newton 1986, Newton, Dale & Rothery 1997, Thompson et al. 1998] Although density-dependency was found in relation to predation by sparrow hawks on the Scoughall site, Whitfield states it is unlikely this is the case on all sites’, where redshanks and sparrow hawks exist.

The sparrow hawk is not the only predator of wintering redshanks, since peregrines, Falco peregrinus, also kill redshank. Peregrines are more likely to chase prey, and choose to kill adults rather than juveniles. Peregrines have less hunting restrictions than the sparrow hawk and are not as regular a predator to redshank as sparrow hawks are. [Kenward 1978, Temple 1987, Whitfield et al. 1999] both raptors have different winter ranging behaviour, which can be affected by severe weather. Although no strong correlations were found between weather and mortality, it was found that wind speed contributed to the influence of density on juvenile mortality. Furthermore, low temperatures can increase energy requirements of prey [Wiersma & Piersma 1994] and predator [Masman, Daan & Belhuis 1988] and sparrow hawk are more successful at catching redshank in low temperatures. [Hilton et al. 1999]

Whitfields report was important in that it identified density-dependent mortality in redshank from sparrow hawk predation. However, it contains no hypothesis for the anomalous winter of 1989. No possible reasons were offered for the unusually high mortality.

Although Whitfield did not consider weather as a strong factor, the winter of 1989 [13 February] is recorded as having the highest gust levels in the British Isles since 1969. The Kinnaird lighthouse recorded a gust of 123 knots [141mph]. The year was also an exceptionally wet one for Scotland. Since the 1989 gust level was the highest ever recorded, any previous studies comparing wind speed on redshank mortality or raptor predation would not have included data on wind gusts at this level. Also, Whitfield reports that wind speed only marginally failed to pass the criterion set for inclusion in the regression model’ for weather variables. Extreme weather conditions could have had some effect on mortality in the winter of 1989, such as sparrow hawk ability to easier take prey. [e.g. – increased risk of predation from sparrow hawks in wind and rain. Hilton, Ruxton & Cresswell 1999] Redshank density was also higher in 1989, a possible effect from the change in the food supply.

Whitfield does, however, mention that hawk attack success was greater in winters with higher redshank density’. Higher density of redshank in 1989 could also have been due to long-term population changes. In wetland bird surveys between 1969 and 1996, Austin, Peachel & Rehfisch found that in south-east Scotland waders showed long-term population changes. Bird numbers other than redshank could have been low, which would have given less optional prey to predators.

Extra predation by peregrine may have accounted for some losses. Although they had declined in 1950/60s from the use of pesticides, numbers increased from 1961. In 1971, 489 pairs were recorded in the British Isles. By 1991 there were 1283 pairs [Ratcliffe 1993] Over 1000 were nesting in 1989. Likewise sparrow hawks have enjoyed a gradual increase in numbers, peaking in 1989/90 and stabilising since. [Newton 1986]

Another factor that may have had an effect on density, which was not considered by Whitfield, was the feeding behaviour of redshank and associated costs. Redshank feed in tight flocks at night and more scattered or solitarily during the day. The difference is attributed to daytime feeding on small shrimps by sight where feeding rate is greater with increased neighbour distance. Feeding close together during the day causes interference. [Goss-Custard 1976] Night time feeding is taken with birds in closer proximity since redshank then trawl in the mud with their beaks for snails. Neighbour disturbance is not a factor in night feeding and therefore the birds flock closely together. This is a feeding behaviour and so flocking together is not only undertaken because of predators.

Although a good report, Whitfield could have considered other factors, which may have explained some of his results more thoroughly.