3.2. Prey choice and timing of breeding

Grouse were clearly the most preferred prey in early spring, especially the smallest species willow grouse and hazel grouse (II). The preference of grouse species and also other species varied markedly between years and within the breeding season (I, II, V). The main prediction of the optimal foraging theory predicts the diets accurately when the prey is immobile (Pyke et al. 1977, Sih 1990). However, in mobile higher vertebrates behaviour and learning changes the vulnerability of prey and the properties of predator (Hughes 1979). For example, the proportion of black grouse hens in the diet has been found to vary depending on the phase of the vole cycle, independent of its abundance in the field (Widen et al. 1987, Selås 1998, V). This is caused, probably, by the change in the behaviour of hens. The theory predicts a preference for large prey if handling times are equal for all preys. This is surely not the case. Changes in preference are clearly affected by changes in handling times. Chicks are easier to catch than adult birds. In spite of the preference shift to small prey in summer, mainly chicks over adult birds, adult grouse maintained their relatively high preference even then (II). The preferred species were not always very important in the nutrition of the goshawks, at least in summer. This might suggest that some prey species were exceptionally vulnerable to predation and were taken every time they were encountered. The weak point in testing OFT with diets is that they are assumed to be the outcome of prey choices. Yet, they are rather the outcome of successful attacks of a predator (Sih 1990, Sih & Moore 1990). Attacks are probably launched much more often, also for species not found in the diet (Cresswell 1995). Preference for prey size was clearest in late summer when goshawks switched to grouse chicks. Grouse broods can be kept as ‘patches’ where the predator can visit several times in succession (Sulkava 1964, Redpath et al. 1997). The predictability of such a patch probably depends much on the structure of the habitat. Grouse broods move mainly after hatching, which may be an anti-predator behaviour against avain predation (Sonerud 1985). If the habitat fragments it would be easier for the predator to locate the ‘patch’ again. This might be the case, at least for capercaillies, the most tightly bound to old forests (Wegge et al. 1990, Storaas et al. 1999). Soon, predator searching time decreases for the prey, which leads to an increase in the profitability.

The availability of suitable prey in early spring is crucial for starting breeding. In raptors, there is a clear division of roles in sexes (Newton 1979). Males provide the food for the female and brood until fledging (Sulkava 1964, Kenward et al. 1993a). During the radio-tracking study I found that females essentially reduced their wintering range and activities were concentrated near the breeding site by the beginning of March (c.f. Ziesemer 1981, Newton 1986). Females rarely moved more than one kilometre from the nest. However, it was not often I found many prey remains in the breeding sites before April. Females were probably still hunting in March. This is based on hare pellets found very often near the nests in early spring. Mountain hare is profitable prey for females due to its large size. After killing a hare, females don’t need to move over wide areas searching for more prey for 3-5 days, which helps raise the body condition for egg production (Widen 1985, Newton 1986, Meijer et al. 1988).

Starting the breeding very early is probably selected for in goshawks (c.f. Lack 1954, Perrins 1970, Newton 1979). The egg laying period overlaps with the start of grouse displaying and leking, when they are more vulnerable to predation (Angelstam 1984, Willebrand 1988, Nielsen & Cade 1992, Valkeajärvi & Ijäs 1994). Starting early ensures that fledglings reach their independence respectively early, which gives them more time to train their hunting skills in relatively favourable conditions before the harsh winter time (II). Better survival rates in early broods have been documented in several raptor species (Marquiss & Newton 1984, Newton 1986, Village 1990). Thus, the timing of breeding in the goshawk supports Perrin’s (1970) theory which is to start breeding as early as possible, in contrast to Lack’s (1954) theory that predicts adjusting the highest demand of offspring in nestling phase to the best supply, which in the goshawk’s case would be a much later date (II). On the other hand, availability of easily catchable nestlings and fledgings of song birds and corvids may be highest during the nestling phase of the goshawk (Toyne 1998). Selection for early breeding can cause difficulties in two other critical phases of the breeding in raptors, namely starting the breeding and hatching the chicks (Newton 1979). By postponing the breeding goshawks would avoid food shortages in early spring and early summer, in particular during cold springs, when grouse can cease leking and females gather in flocks (Marjakangas 1986, Elkins 1988, Nielsen & Cade 1992), which makes hunting them more difficult. At the hatching phase, the food demand of the hawk family sharply increases (Tolonen 1994, Kennedy & Ward 1994). In late springs males may have difficulties providing enough food, when the growth of vegetation impairs the perceptiveness of ground dwelling prey. Goshawks rely a lot on thrushes, corvids and their nestlings in this phase. (Linden & Wikman 1983, II, S. Sulkava, pers. comm.). If prey, even small ones, are not delivered to the nes often enough the weakest chick or chicks will easily die. It is at this time that brood reduction in goshawk is most common (Huhtala & Sulkava 1981).