29. COULD THE HEN HARRIER (CIRCUS CYANEUS) DECLINE ON ORKNEY BE DUE TO A SHORTAGE

Transcription

1 29. COULD THE HEN HARRIER (CIRCUS CYANEUS) DECLINE ON ORKNEY BE DUE TO A SHORTAGE OF FOOD? Arjun Amar, Steve Redpath, Xavier Lambin & Eric Meek Summary 1. The population of hen harriers Circus cyaneus in Orkney has declined dramatically over the last 20 years. One hypothesis for this decline is that there has been a reduction in the amount of food available. 2. Harriers in Orkney were found to hunt areas with more rough grass during both the spring and summer. 3. Agricultural records revealed that, since 1960, there has been an increase in the proportion of land under pasture and a decrease in the amount of rough grazing. Additionally, since the start of the 1980s, stocking densities of sheep Ovis aries have more than doubled on mainland Orkney. These changes are likely to have been detrimental to harriers and could have reduced the amount of food available. 4. Male harriers in Orkney provisioned food to their females during the pre-lay period at a lower rate than did males in another Scottish population, where the population was breeding far more successfully. 5. A supplementary feeding experiment showed that males provided with extra food had more breeding females than control males. 6. Observational and experimental data therefore suggested that the population is indeed limited by food and that a reduction in prey is a likely explanation for this population s decline Introduction The Orkney hen harrier Circus cyaneus population was once the stronghold for this species in Britain, but the population is currently in decline (Meek et al., 1998). The hen harrier became extinct as a breeding species on mainland Britain at the start of the 1900s, due to human persecution, and it is believed that birds from the Orkney population aided in the re-colonisation of the British mainland (Watson, 1977). Hen harriers on parts of mainland Britain are persecuted because they kill red grouse Lagopus lagopus scoticus and can limit the numbers available for shooting in autumn (Thirgood et al., 2000; Redpath & Thirgood, Chapter 39). Persecution of this species continues despite the fact they are legally protected under international and UK legislation (e.g. Etheridge et al., 1997; Stroud, Chapter 3). However, in Orkney persecution does not occur due to the absence of commercial grouse shooting. The decline of the hen harrier in Orkney is therefore alarming, and if levels of persecution on mainland Britain were to increase, this population may no longer be able to act as a sustainable refuge in the future. 377

2 Birds of Prey in a Changing Environment Hen harriers in Orkney have been monitored with varying degrees of intensity for the last 50 years (Balfour, 1957; Balfour & Cadbury, 1979; Picozzi, 1984a; Downing, 1990; Meek et al., 1998), and this continuous effort has allowed this population s decline to be revealed. Meek et al. (1998) found that numbers of males and females had declined by around 70% and that productivity had decreased by 60%, from the peak figures of the 1970s. During previous research, male hen harriers in Orkney were found to be highly polygynous (Balfour & Cadbury, 1979; Picozzi, 1984a,b). Both authors considered that the female biased sex ratio of the population was an important factor in causing this high level of polygyny. One of the major features associated with the decline of this population has been the reduction in polygynous breeding, defined as a male mated with more than one female, which lays at least one egg. Picozzi (1984a) found that from , an average of 77% of adult males were polygynous breeders. Current levels of polygyny in Orkney are far lower (only 17% in ) and some males fail to breed at all (Amar et al., in press). Despite the fact that polygynous breeding is low there is still a female biased sex ratio in the population. The female:male sex ratio in 1998 was estimated to be 1.8:1 in Orkney s West Mainland, which is similar to the average bias found during , which was 2.2:1 (Picozzi, 1984a). These figures suggest that a shortage of males is not the only factor involved in influencing levels of polygyny on Orkney. A shortage of food during the early breeding period may be responsible for the reduced levels of polygyny and also for the poor productivity of the population, factors that may ultimately be responsible for this population s decline. Polygynous breeding is often associated with productive habitats (Verner & Willson, 1966). If levels of food have decreased in Orkney, males may not be able to supply adequate food to allow more than one female to breed. Changes in land use, such as the destruction of habitats and the intensification of farming, may be factors responsible for the shortage of prey. In this chapter we explore the hypothesis that a shortage of food during the early breeding period could be responsible for the population s decline. We examine the changes in land use on Orkney and investigate the implications of these changes for harriers by examining which habitats and habitat characteristics are important to hunting harriers. We compare the provisioning rates of male harriers in Orkney with the rates from males at Langholm, in south-west Scotland, where the population breeds more successfully (Amar et al., 2003), in order to examine whether differences in breeding performance are reflected in provisioning rates. Finally, we examine the effect of a supplementary feeding experiment on the breeding performance of the hen harriers in Orkney Methods Changes in land use In Orkney, the following parishes in the West Mainland have breeding hen harriers: Birsay, Harry, Evie, Rendall, Firth, Ophir and Stenness. Data from the June Agricultural Census from 1960 until 1998 in these parishes were used to assess changes in land use. Data were collected on the total area under pasture and under rough grazing, the total area of land from which records were taken in each parish and the total number of sheep Ovis aries in each parish. These latter two measures were then used to calculate the number of sheep per hectare for each parish in each year. 378

3 Could the hen harrier decline on Orkney be due to a shortage of food? Habitat use by hunting harriers The relative use of three habitat types by hunting harriers was investigated. These habitats were as follows: i) upland moorland dominated by Calluna vulgaris; ii) semi-natural lowland principally a mixture of Calluna vulgaris and rough grass, similar to the transitional land between moorland and farmland; and iii) intensive lowland farmland dominated by sheep and cattle Bos taurus pasture, but also with small areas of semi-natural lowland, primarily rough grazing. Squares of 1 km dominated by each habitat types were chosen subjectively based on appearance. Although ideally squares would have been chosen randomly, the limited areas of certain habitat types (principally the semi-natural lowland) combined with suitable topography to achieve sufficiently good observations meant that these factors constrained selection of squares. However, as we had no previous knowledge of harrier hunting intensity, potential bias in site selection was unlikely. No squares had active harrier nests within them during a watch period and all squares were located within 5 km of an active harrier nest. In summer, a nest proximity index for each square was then calculated as the sum of the reciprocals of the squared distance to each nest site. In spring, a similar territory proximity index was calculated, using the same methods, but replacing distance to nests with the distance to the centre of male territories in the subsequent breeding season. Data on habitat use by harriers were collected in summer 1998, spring and summer 1999 and spring Data were collected in spring between 28 February and 29 March and during summer from 12 July until 16 August. Nine squares, three of each habitat type, were used in all periods, except for the three moorland squares in summer 1999 (due to the presence of active harrier nests). Data in summer 1999 and spring 2000 were also recorded from a further nine squares, again, three in each habitat type. Data were therefore collected from a total of 18 squares. Watches of squares were conducted from suitable locations that were close to the edge of the square and also offered good views over the whole square. The relatively flat landscape of mainland Orkney allowed continuous observations of harriers within a square. Watches were conducted at squares until a total of ten hours watching per square had been accumulated (usually in two to three hour watch blocks). During a watch, squares were continuously scanned using 10 x 42 binoculars for hunting harriers. Harriers were classified as hunting if they were flying approximately <10 m from the ground (estimated visually). Harriers tend to hunt by flying low and quartering across the ground (Schipper et al., 1975) and <10 m was the criterion chosen to exclude harriers which may not have been actively searching for prey. Previous work by Madders (1997) showed that there was a high degree of agreement between two observers, independently recording the height of an individual harrier and that this close level of agreement also existed between habitat types. A tape recorder, set to run continuously for the duration of each foraging bout, was used to record the duration of hunting within a square. Vegetation composition and height were recorded between the same dates as the hunting watches were conducted and were performed in both spring and summer. For squares observed during the same season in two years, vegetation measures were only recorded in the first year as vegetation was assumed not to have altered greatly between years. Quadrats, measuring 25 cm 2 were placed every 40 m along two parallel transect lines within a square, each located 250 m and 750 m away from one side of a watch square, thus giving 379

4 Birds of Prey in a Changing Environment information from 50 quadrats per square. The dominant type of vegetation in each quadrat was recorded and the highest live point of this vegetation type measured. Vegetation was recorded at species level, except for grass species, which were classified categorically as either smooth or rough grass. A build up of dead vegetation adequate to conceal a moving vole (Hewson, 1982) defined rough grass. Habitat use by hunting harriers was examined in relation to the three main habitat types in the two seasons. At a finer scale we examined the amount of hunting in an area in relation to the abundance of the main vegetation types (i.e. those that were dominant in over 10% of the quadrats in a squares), the mean vegetation height and the proximity of nests or territories in spring and summer, respectively Comparative breeding success and provisioning rates Possible harrier territories in Orkney and Langholm were checked at the start of the breeding season for signs of occupancy, which was identified from either perched or displaying males or females. During this study some birds in Langholm and Orkney were provided with supplementary food (see later, but see also Redpath et al., 2001; Amar & Redpath, 2002). All data relating to birds provided with supplementary food were excluded from these analyses, as was information relating to the second breeding attempt of birds which re-laid following the failure of their initial clutches. The rate at which males provisioned their females with food in Orkney and Langholm was recorded during both the pre-lay and incubation period. The pre-lay period was defined as four weeks prior to laying, or if no clutch was produced, until the mean laying date for that year. Watches were conducted at times with good visibility and little precipitation. In Orkney, an effort was made after a food pass to search for prey remains in the area where the female fed and the prey types were identified to provide information on the diet of the birds in this area. Similar data were not collected in Langholm, but observations indicated that all items delivered were small, probably almost exclusively small mammals and meadow pipits Anthus pratensis or other small passerines (Redpath & Thirgood, 1997). So as to investigate whether there were large differences in the average sizes of prey between the areas, we estimated the average biomass of prey items delivered in the two localities. In order to calculate these we used the percentage of prey types identified from remains in Orkney, and for Langholm we assumed an equal representation in the diet of field voles Microtus agrestis and passerines. However varying this assumption for Langholm would have little effect as passerines and field voles weights only differed by 13 g. Rabbit Oryctolagus cuniculus biomass was calculated from the regression equation from Carss (1995), using the mean hind foot length of rabbits found in the nest during a previous study on Orkney (Picozzi, 1980). A quarter of this biomass was used to account for the fact that only part of a rabbit was usually delivered by the males. Biomass of Orkney vole Microtus arvalis orcadensis or field vole was taken from Corbett & Harris (1991) and the meadow pipit s weight from Picozzi (1978) was used for passerines Supplementary feeding experiment A short-term trial in 1998 to provide supplementary food to hen harriers in Orkney found that scavenging birds, particularly hooded crows Corvus corone corone, quickly removed all the food intended for the harriers. So as to achieve supplementary feeding of the harriers it 380

5 Could the hen harrier decline on Orkney be due to a shortage of food? was therefore necessary to remove crows from the immediate areas. This, however, potentially confounded the effect of feeding and therefore an additional group from which crows were removed was also included in the experiment. This additional removal group therefore allowed assessment of the effect of crow removal independently of food provision. If the removal of crows had no effect on breeding parameters, we could then realistically conclude that any effect in the fed group was most likely attributable to the provision of food rather than from the removal of crows. We also had a third control group where neither crow removal nor supplementary feeding occurred. Our experiment therefore had three treatments, namely:- 1. Fed group, where supplementary food was provided and crows were removed; 2. Removal group, where crows were removed from harrier territories; and 3. Control group, where neither crow removal nor supplementary feeding occurred. The experiment was conducted during 1999 and 2000, but we also used data from unmanipulated nests in 1998 in order to increase sample size of the control group and to test for territory effects. Territories used by harriers in 1998 were randomly assigned to one of the three treatments in 1999 before the start of the breeding season. Subsequently in 2000, territories were randomly assigned to one of the other two treatments. Supplementary food was placed on feeding posts and at likely perching places from 6 April and continued throughout the incubation period in both years. Supplementary food mainly consisted of dead day old cockerel chicks but also some pieces of rabbit and hare Lepus europaeus. For full methods of the supplementary feeding experiment see Amar & Redpath (2002). All harrier territories were watched throughout the pre-lay and laying period both to count numbers of females associated with males and to find nests. Males were not individually marked, so synchronous watches by multiple observers were performed in adjacent valleys, sometimes with the aid of short-wave radios. Using overlapping and sequential sightings of males, we were confident that we were able to ascertain which males were mated to which females. The maximum number of females associated with each male during the pre-lay period was based on courtship flights, birds perched together, food passes and nest building. The pre-lay period was defined separately for each year as the period from the start of April until the median lay date for that year. This period was used to account for the many females that did not lay and so allowed this period to be standardised between all males. In this chapter, breeding females were defined as those that were confirmed to have laid at least one egg. The breeding status of a female was classified as either monogamous, primary or secondary (the last category including both beta (secondary) and gamma (tertiary) females). Polygynous females were classified according to lay date, with the earliest laying female being considered the primary female. Nests were usually located during egg laying or early incubation, and both lay date (first egg) and clutch size recorded. For full methods on how lay date was calculated see Amar & Redpath (2002). Nests that had either uncertain lay date or clutch size were excluded from any analysis relating to these measures Statistical analysis All statistical analyses were performed using SAS, version 6.12 (SAS Institute, 1990). Generalised linear mixed models were implemented using the GLIMMIX macro (Littell et 381

6 Birds of Prey in a Changing Environment al., 1996) and generalised linear models, using the GENMOD procedure. Data on habitat use were analysed using a gamma error structure and a log link function. The properties of the gamma distribution mean that zero figures cannot be accommodated; to accommodate this we added half of the smallest value of the response variable being examined to each observation. We chose this value because the analyses were conducted using a log link function; values at the lower end of the range would therefore have had a disproportionate effect in the analyses. Because some squares had repeat observations in the different years, we analysed the data using GLIMMIX with the non-independent effect of individual squares set as a random term in the model. There were, however, insufficient replicates in the data set to examine the summer data using GLIMMIX; so instead GENMOD was used. To accommodate the potential pseudo-replication from repeated observations on the same sites, squares with repeat observations carried half of the weight in the analysis. Provisioning rates of male harriers were analysed using GENMOD, with a Poisson error structure and a log link function. The effect of experimental treatment on breeding parameters was analysed using GLIMMIX, with the non-independent effects of the different male territories incorporated into the analysis as a random term in the model. In this analysis data from un-manipulated nests in 1998 were also used to increase sample size and to further control for territory effects. Count data such as the maximum number of females associated per male and the number of breeding females per male were analysed with a Poisson error structure and a log link function. Lay date and clutch size were analysed with a normal error structure, and an identity link function and hatching success, a binary measure, was analysed with a binomial error structure and a logit link function. All models had year and treatment as permanent fixed effects because 1998 lacked any data for fed and removal territories. For models analysing breeding parameters determined at a female level, such as lay date, clutch size or hatching success, we also tested for the effect of female breeding status. All models were constructed using a backward elimination procedure, dropping the least significant term in the subsequent model until only terms significant at the 10% level, using type III analyses, remained. All pair-wise comparisons were conducted in the GLIMMIX models, through t-tests of the differences in least square means (DLSM) and in generalised linear models using specific contrast statements. For the models implemented using the GLIMMIX procedure, denominator degrees of freedom were estimated, using Satterthwaite s formula (Littell et al., 1996) Results Change in land use Figure 29.1 shows the mean changes in land use and sheep stocking density from the seven parishes in West Mainland that were investigated. Since 1960 there has been an increase in the amount of land under pasture from 36% to 54% in Over this same period there has been a corresponding decline in the amount of land under rough grazing by 11%, equivalent to a loss of 22 km 2. Over the same period, numbers of sheep have changed dramatically. Numbers remained relatively constant between 1960 and 1980 but they then increased sharply from 1981 onwards. During the period the mean stocking density was 0.95 sheep per hectare, by 1998 stocking densities were over twice this level at 2.14 sheep per hectare. 382

7 Could the hen harrier decline on Orkney be due to a shortage of food? Figure Proportion of land under pasture (pale line) and rough grazing (solid line) and the density of sheep (dashed line) in Orkney from Data are mean values from seven parishes on West Mainland Orkney which have breeding hen harriers Habitat use by hunting harriers The total time for which harriers were observed hunting in the squares did not vary significantly between spring and summer (F 1,33 = 1.20, P = 0.10). During the spring there was a positive association between harrier hunting and territory proximity (F 1,21 = 6.47, P <0.02). In other words, squares were generally hunted more the closer they were to the eventual breeding territories. Therefore, in subsequent analyses examining hunting in spring, territory proximity was controlled for by having it as a permanent fixed effect in the models. After controlling for this variable, the habitat type in which squares were located did not significantly explain the variation in the amount of time that squares were hunted (F 2,17 = 0.77, P >0.40). Examining the amount of time that squares were hunted in relation to the vegetation within the squares revealed that, in spring, there was a near-significant positive association with the amount of quadrats dominated by rough grass (F 1,16 = 3.68, P <0.08), once the proximity of territories were controlled for. In summer, there was no effect of nest site proximity on the amount of time that a square was hunted. However, the habitat in which squares were located did explain a significant amount of the variation in hunting time between the squares (F 2,21 = 4.54, P <0.05), with semi-natural lowland habitats being hunted significantly more than either farmland (F 1,21 = 7.00, P <0.05) or moorland (F 1,21 = 4.62, P <0.05) (Figure 29.2.). In summer, the amount of time that an area was hunted by harriers was significantly positively associated with the amount of rough grass (F 1,21 = 12.5, P <0.01) after controlling for the nearsignificant effect of heather cover (F 1,21 = 3.34, P <0.10). 383

8 Birds of Prey in a Changing Environment Figure Amount of time harriers were seen hunting per ten hours in 1 km squares in the three habitat types in spring (clear bars) and summer (filled bars). Sample size for each bar is nine, except moorland in summer where sample size is six. Data are given as means with one standard error Breeding success and male provisioning rates in Orkney and Langholm In the Langholm study area, between 1992 and 2000, 100% of the females that settled on a territory subsequently laid a clutch. The situation in Orkney was remarkably different. Between 1998 and 2000, over 50% of females that were on territories failed to lay a clutch. Hatching success also differed between Orkney and Langholm, although the difference was less striking. In Langholm, between 1993 and 2000, 13% of females failed to hatch their clutch, whereas in Orkney, between , 35% of breeding females were unsuccessful at hatching their clutch. The frequency with which prey items were delivered to females by male harriers in Langholm was significantly higher than in Orkney, during both the pre-lay (F 1,44 = 69.7, P <0.001) and incubation periods (F 1,46 = 19.3, P <0.001) (Figure 29.3). Between the prelay and incubation periods, there was an approximate four-fold increase in the rate at which food was delivered by males in Orkney and this increase was statistically significant (F 1,56 = 14.69, P <0.001). In Langholm, however, provisioning rates did not vary between these two periods (F 1,34 = 0.19, P >0.50). Remains from 13 prey items were found after food passes during the pre-lay and incubation periods in Orkney (Table 29.1a). From the remains of young rabbits found, it was evident that males usually only delivered a part of the rabbit, often only a leg or a head. After calculating the average biomass of prey items it did appear that males in Orkney tended to provide more biomass per prey delivery. We calculated an average biomass per prey item of 47.6 g and 23.7 g for Orkney and Langholm, respectively. If the respective provisioning rates in the two localities are multiplied by the average biomass of prey items, we can derive the approximate average biomass (g) delivered per hour 384

9 Could the hen harrier decline on Orkney be due to a shortage of food? Figure Provisioning rate (items delivered per hour) of male harriers in Orkney (filled bars) and Langholm (clear bars) during the pre-lay and incubation periods. Data given as means with one standard error. (Table 29.1b). After accounting for these biomass differences, provisioning rates during the pre-lay period appeared to be around six times higher in Langholm than in Orkney. However the differences between the two locations during the incubation period appears to be less marked. Therefore, during the incubation period, there is a possibility that although items are delivered less frequently in Orkney, the amount of food delivered does not differ greatly. Table (a) Prey items identified from sites where females fed after food passes, during the pre-lay and incubation periods in Orkney, 1998, 1999 and Sample size of prey types in parentheses. (b) Estimation of the average biomass (g) delivered per hour in Orkney and Langholm during the pre-lay and incubation periods. Data are derived from the mean provisioning rates in the respective areas and the average biomass of prey delivered (see text for more details). (a) Prey type Percentage Biomass (g) Young rabbit 38 (5) 277 Orkney vole 38 (5) 39 Passerine 23 (3) 17 (b) Study site Provisioning rate (g per hour) Orkney Langholm Pre-lay Incubation

10 Birds of Prey in a Changing Environment Figure a) Mean number of breeding females shown as clutches produced per male in each of the treatment groups for all years combined. N = number of males. b) Probability of a female hatching a clutch in relation to which treatment their male was receiving for all years combined. Data are presented as means with one standard error and sample size. N = number of females Supplementary feeding experiment The number of females associated with males during the pre-lay period, varied significantly between years (F 2,29 = 10.12, P <0.001). Although, experimental treatment had no significant effect on the number of females associated with males (F 2,32 = 2.43, P >0.10), it did have a significant effect on the number of females breeding per male (F 2,31 = 7.71, P <0.005) (Figure 29.4a), with fed males having significantly more breeding females than controls (DLSM: P = 0.003). There were no differences in the number of females breeding per male between the removal and control groups (DLSM: P = 0.87). There was also a significant effect of year (F 2,28 = 3.69, P <0.05), with fewer breeding females per male in 1999 compared with either 1998 (DLSM: P <0.01) or 2000 (DLSM: P <0.05). There was no significant effect of treatment on lay date (F 2,33 = 0.46, P >0.50), nor was there any effect of year (F 2,27 = 1.68, P >0.10). There was, however, a near-significant effect of female breeding status (F 2,32 = 3.22, P <0.10), with primary females laying significantly earlier than monogamous females (DLSM: P = 0.02) and there was also a tendency to lay somewhat earlier than secondary females (DLSM: P <0.10). So neither the provision of supplementary food nor the removal of crows significantly affected lay date. There was no significant effect of treatment (F 2,30 = 0.21, P >0.80), year (F 2,24 = 0.50, P >0.60) or breeding status (F 2,29 = 1.42, P >0.20) on clutch size. Hatching success was also unaffected by treatment (F 2,38 = 0.87, P >0.40) (Figure 29.4b). However, there was a significant effect of year on hatching success (F 2,39 = 3.43, P <0.05), with clutches more likely to hatch in 2000 than in 1999 (DLSM: P = 0.01). Breeding status of the female also significantly affected hatching success, with secondary females being significantly less likely to hatch their clutch than either primary (DLSM: P <0.01) or monogamous females (DLSM: P <0.005). There were insufficient data to consider the interaction between female status and treatment on hatching success in the GLIMMIX model. However, if the hatching success of secondary females, mated with fed and control males, was pooled between all years, we found that no secondary females mated with controls were successful (n = 5), whereas 66% of secondary females mated to males supplied with food were successful (n = 6), a difference that was found to be statistically significant (Fisher exact test P = 0.045). 386

11 Could the hen harrier decline on Orkney be due to a shortage of food? 29.4 Discussion Large-scale changes in land use in Orkney have occurred since Based on the pattern of habitat use by hunting harriers, these changes are likely to have been detrimental to harriers in Orkney. The decrease in the amount of rough grazing and the increase in sheep numbers will have reduced the amount of rough grass in the environment, which is the favoured habitat for hunting harriers in the spring and the summer. The loss of rough grazing habitat and the increase in the amount of intensive pasture may have resulted in a shortage of food for the harriers, as voles and meadow pipits are more abundant in moorland and semi-natural lowland habitat than they are in intensive farmland. Indeed vole and meadow pipit abundance are known to correlate with the amount of rough grassland in some of these habitat types (Amar, 2001). Observational data supported the idea that there was a shortage of food for the harriers in Orkney. Males in Orkney delivered food less frequently during both the pre-lay and incubation periods than males in Langholm, where the population breeds more successfully. After accounting for the differences in prey biomass between these populations, there was a suggestion that it was only during the pre-lay period that these large differences remained. However, young rabbits are relatively large prey items and formed 38% of the prey remains found in Orkney and were therefore largely responsible for harriers in Orkney receiving more biomass per prey delivery. Their importance might, however, be overestimated, since remains of rabbits may be easier to find than smaller items, such as passerines or small mammals, which may be completely eaten (Picozzi, 1980). However, the average biomass of prey delivered by males during the nestling period, extrapolated from Reynolds (1992) and Picozzi (1980), is 57.1 g and is therefore not particularly different from the estimate of 47.6 g used in this study. It is not known to what degree the size of prey items varies between individuals within Orkney, which therefore makes any statistical testing of biomass provisioned problematic. It appears, therefore, that the discrepancy in provision rates between the two populations may be most acute during the pre-lay period. This is especially interesting, since the most striking difference in breeding success between these two populations was in the proportion of settled females which subsequently bred. In Langholm, all females that were on territory in spring went on to lay a clutch, suggesting that males were able to provide them with adequate food during the pre-lay period. This contrasts sharply with Orkney, where 50% of the females failed to lay. The large differences in pre-lay provisioning rates between these two populations may well explain these differences in breeding success and indicates that the breeding success in Orkney could have been reduced as a result of a shortage in food. Another interesting feature in the comparison between these two populations was that, in Orkney, provisioning rates during the incubation period were significantly higher than during the pre-lay period, whereas in Langholm no such difference was found. This further suggests that the shortage of food for the harriers in Orkney may be particularly acute during the earlier stages of breeding. It therefore appears that the rate at which food is supplied to the females during the prelay period may have declined over the last 20 years and that this may be responsible for the reduced breeding success currently seen in the population. Unfortunately there is no quantitative data available from earlier times which would allow this assumption to be tested. Interestingly, Nick Picozzi, who studied this population during the late 1970s, 387

12 Birds of Prey in a Changing Environment considered that food provisioning was never prolific in this population (N. Picozzi, pers. comm.). If provisioning rates have not changed dramatically then an alternative explanation, evoking the same ultimate factor of food shortage, might be that the females are currently in a poorer condition prior to the period of pre-lay courtship feeding due to the overall shortage of food in the environment. In that case the rate at which prey is delivered, although not different from previously when the population bred more successfully, is now no longer sufficient to allow as many females to breed. Regardless of which scenario is actually occurring, the ultimate cause of the reduced breeding success is still a shortage of food available in the environment. The provision of supplementary food to male hen harriers increased their number of breeding females, and therefore provided the necessary experimental evidence to show that breeding performance in this population is currently limited by food. The increase in breeding females resulted from a reduction in the number of males failing to breed and also an increase in the amount of polygynous breeding. All fed males were mated with at least one breeding female, whereas around 20% of both controls and removal males had no breeding females. Over the two experimental years, 36% of fed males bred polygynously, including two cases of trigamy. In the other two groups there was only one case of polygyny, representing only 5% of un-fed males. Territorial polygyny in birds is rare (Lack, 1968; Møller, 1986). Where it does occur, it is often associated with areas or years of high food abundance (Verner & Willson, 1966; Hamerstrom et al., 1985; Korpimäki, 1989). For species in which breeding females are dependent on males for food, the association between territorial polygyny and food is often thought to result from conditions that allow males to support more than one breeding female (Orians, 1969). In our study, similar numbers of females were associated with fed males as were associated with males in the other two experimental groups, yet fed males bred with significantly more females. This suggests that they were able to provision food at a sufficient level to allow more of their associated females to breed. These results are consistent with findings from correlative studies of northern harrier populations (Hamerstrom et al., 1985; Simmons et al., 1986), whose main prey, Microtus voles, undergo large cyclic fluctuations in abundance. The highest levels of harrier polygyny were associated with years of high Microtus abundance. Simmons et al. (1986) suggested that polygynous breeding by northern harriers was only possible when males were able to supply food to secondary females above a certain threshold rate during the pre-lay period. Previous studies of the Orkney population found that, on average, around 75% of adult males were polygynous (Balfour & Cadbury, 1979; Picozzi 1984a) whereas current levels are far lower. That supplementary feeding increased polygynous breeding therefore suggests that in Orkney there is currently insufficient food available to harriers during the pre-lay period to allow many of the polygynous associations to be expressed as polygynous breeding. Food had no significant effect on lay date or clutch size. We believe this was related to the time that males first started to take the supplementary food. Only 50% of all females breeding with fed males, for which we had accurate lay dates, were paired with males that were seen taking food before laying commenced. So food levels for many of the fed males may not have been effectively increased until after laying had started. However, in all cases of secondary females with accurate lay dates, males were seen to take food before laying 388

13 Could the hen harrier decline on Orkney be due to a shortage of food? had commenced. This seems to be a likely explanation for the observation that supplementary feeding caused fed males to have more breeding females yet had no significant effect on lay date or clutch size. In a supplementary feeding experiment conducted on the hen harriers in Langholm, Scotland (Redpath et al., 2001), most males were seen to take supplementary food before laying commenced (S. Redpath, pers. obs.). This experiment again found no significant effect on lay date, but supplementary feeding did significantly increase clutch size (Redpath et al., 2001). The difference in the effect of feeding on clutch size between the two experiments may be explained by the fact that males started taking food relatively earlier in the experiment at Langholm. Another interesting difference between these two experiments was that, unlike in Orkney, feeding males in Langholm did not increase the number of breeding females. This reinforces the suggestion that breeding success in the two areas is influenced by the rate at which males provisioned food. Males in Langholm are able to provision food at an adequate rate, and so may not be as limited by food. The lack of effect of supplementary food on hatching success was surprising, but could perhaps be due to the same reasons. If food levels were not effectively increased for many of the females before laying commenced, then their condition may also not have been increased before laying. Newton (1986) found that female sparrowhawks Accipiter nisus in better condition before incubation were more likely to successfully hatch their clutch. This may be the case with Orkney hen harriers. This possibility is supported by the fact that hatching success of secondary females was significantly higher if they were mated with fed males rather than with control males. As all fed males with secondary females were seen to take food before their secondary females laid, the condition of these females may have been higher before laying than the controls or the primary females of the fed males. The main experimental finding was that the number of females breeding per male was food limited. It should be noted, however, that because crows were also removed on fed territories, the additive effect of supplementary food and decreased predation could not be completely eliminated as a potential cause of this result (see also Amar & Burthe, 2001). However, the fact that crow removal had no effect on any of the breeding parameters in the removal group suggests that crow removal was of little overall consequence and was therefore unlikely to have had a strong confounding effect in the fed group. In conclusion, both comparative and experimental data supported the idea that the hen harrier population in Orkney is currently limited by food availability during the early part of the breeding season. A reduction in the prey available to harriers is therefore the most likely explanation for this population s decline. This reduction in food may be attributable to a decrease in the amount of rough grazing which harriers selectively hunted in the spring and summer, and the increase in sheep stocking density which is also likely to have decreased the amount of rough grass and therefore also potentially the amount of food. Management for this population should be directed toward increasing the amount of rough grass and rough grazing (Scottish Natural Heritage, 2003). This management might be best achieved through the exclusion or reduction of grazing within the farmland areas close to the moorland sites where the harriers breed, and a decrease in the grazing pressures in areas of rough grazing which are currently overgrazed. In January 2003, Scottish Natural Heritage launched a land management incentive scheme (Orkney Moorlands Natural Care Scheme) to create more such rough grassland habitat for harriers. 389

14 Birds of Prey in a Changing Environment Acknowledgements We are grateful to Brian Ribbands, Kerry Lock, Rory Gordon, Sarah Burthe, Gaetan Bottin, Maxime Esnault, Jacqui Todd, Keith Fairclough and Andy Knight for their tremendous help with fieldwork. Special thanks are also due to Donnie Littlejohn for making the crow traps. We would also like to thank the landowners in Orkney for permission to work on their land. Advice on many aspects was gratefully received from Ron Summers, Mick Marquiss, Ian Bainbridge, Audun Stien, Steve Albon, Beatriz Arroyo, Isla Graham, Nick Picozzi and two anonymous referees. Special thanks are also due to Dave Elston for advice relating to experimental design and statistics. This research was funded through an Aberdeen Research Consortium studentship to Arjun Amar and an RSPB grant to Steve Redpath. References Amar, A. (2001). Determining the cause of the hen harrier decline on Orkney. PhD thesis. University of Aberdeen, Aberdeen. Amar, A. & Burthe, S. (2001). Observations of predation of hen harrier nestlings by hooded crows in Orkney. Scottish Birds, 22, Amar, A. & Redpath, S. (2002). Determining the cause of the hen harrier decline on the Orkney Islands: an experimental test of two hypotheses. Animal Conservation, 5, Amar, A., Redpath, S. & Thirgood, S. (2003). Evidence for food limitation in the declining hen harrier population on the Orkney Islands, Scotland. Biological Conservation, 111, Amar, A., Picozzi, M., Meek, E.R., Redpath, S.M. & Lambin, X. (in press). Decline of the Orkney Hen Harrier Circus cyaneus population: do changes to demographic parameters and mating system fit a declining food hypothesis? Bird Study. Balfour, E. (1957). Observations on the breeding biology of the hen harrier in Orkney. II. Bird Notes, 27, Balfour, E. & Cadbury, C.J. (1979). Polygyny, spacing and sex ratio among hen harrier Circus cyaneus in Orkney, Scotland. Ornis Scandinavica, 10, Carss, D.N. (1995). Prey brought home by two domestic cats (Felis catus) in northern Scotland. Journal of Zoology (London), 237, Corbett, G.B. & Harris, S. (1991). The Handbook of British Mammals. Third Edition. Blackwell Scientific Publications. Oxford. Downing, R. (1990). Orkney hen harriers In Orkney Bird Report 1989, ed. by C. Booth, M. Cuthbert & E.R. Meek. Orcadian, Kirkwall. pp Etheridge, B., Summers, R.W. & Green, R.E. (1997). The effects of illegal killing and destruction of nests by humans on the population dynamics of the hen harrier Circus cyaneus in Scotland. Journal of Applied Ecology, 34, Hamerstrom, F., Hamerstrom, F.N. & Burke, C.J. (1985). Effect of voles on mating systems in a Central Wisconsin population of harriers. Wilson Bulletin, 97, Hewson. R. (1982). The effect upon field vole (Microtus agrestis) habitat on removing sheep from moorland in west Scotland. Journal of Zoology (London), 197, Korpimäki, E. (1989). Mating system and mate choice of Tengmalm s Owls Aegolius funereus. Ibis, 131, Lack, D. (1968). Ecological Adaptations for Breeding in Birds. Methuen, London. Littell, R.C., Milliken, G.A., Stroup, W.W. & Wolfinger, R.D. (1996). SAS System for Mixed Models. SAS Institute Inc., Cary, NC. 390

15 Could the hen harrier decline on Orkney be due to a shortage of food? Madders, M. (1997). The effect of forestry on hen harriers Circus cyaneus. PhD thesis. University of Glasgow, Glasgow. Meek, E.R., Rebecca, G.W., Ribbands, B. & Fairclough, K. (1998). Orkney hen harriers: a major population decline in the absence of persecution. Scottish Birds, 19, Møller, A.P. (1986). Mating systems among European passerines: a review. Ibis, 124, Newton, I. (1986). The Sparrowhawk. T. & A.D. Poyser, Calton. Orians, G.H. (1969). On the evolution of mating systems in birds and mammals. American Naturalist, 103, Picozzi, N. (1978). Dispersion, breeding and prey of hen harrier Circus cyaneus in Glen Dye, Kincardineshire. Ibis, 120, Picozzi, N. (1980). Food, growth, survival and sex ratio of nestling hen harriers Circus c. cyaneus in Orkney. Ornis Scandinavica, 11, Picozzi, N. (1984a). Sex ratio, survival and territorial behaviour of polygynous hen harriers Circus c. cyaneus in Orkney. Ibis, 126, Picozzi, N. (1984b). Breeding biology of polygynous hen harriers Circus c. cyaneus in Orkney. Ornis Scandinavica, 15, Reynolds, P. (1992). The impact of changes in land-use in Orkney, on the vole Microtus arvalis orcadensis and its avian predators. PhD. thesis, University of Aberdeen, Aberdeen. Redpath, S.M. & Thirgood, S.J. (1997). Birds of Prey and Red Grouse. The Stationery Office, London. Redpath, S.M., Thirgood, S.J. & Leckie, F. (2001). Does supplementary feeding reduce predation of red grouse by hen harriers? Journal of Applied Ecology, 37, SAS Institute (1990). SAS/STAT Users Guide, Version 6. SAS Institute Inc., Cary, NC. Schipper, W.J.A., Buurma. L.S. & Bossenbrock, P. (1975). Comparative study of hunting behaviour of wintering hen harriers Circus cyaneus and marsh harriers Circus aeruginosus. Ardea, 63, Scottish Natural Heritage (2003). The Orkney Vole A Management Guide. Unpublished report. Scottish Natural Heritage, Perth. Simmons, R., Barnard, P., MacWhirter, B. & Hansen G.L. (1986). The influence of microtines on polygyny, productivity, age, and provisioning of breeding northern harriers: a 5-year study. Canadian Journal of Zoology, 64, Thirgood, S.J., Redpath, S.M., Rothery, P. & Aebischer, N.J. (2000). Raptor predation and population limitation in red grouse. Journal of Animal Ecology, 69, Verner, J. & Willson, M.F. (1966). The influence of habitat on mating systems in North American passerine birds. Ecology, 47, Watson, D. (1977). The Hen Harrier. T. & A.D. Poyser. Berkhamsted. 391

16 Birds of Prey in a Changing Environment 392