The taphonomy of medium-sized grouse in food remains of the northern goshawk Accipiter gentilis, compared with damage done by man and other predators

3.1 Preservation and fragmentation of bones

Medium-sized grouse (L. tetrix and Lagopus spp.) were by far the most numerous victims of A. gentilis; their remains accounted for more than half of all the bones (981 out of 1818) (Table 1). Ducks, corvids, pigeons, large and small grouse (capercaillie and hazel grouse) and hares and squirrels were also relatively numerous. Among the other victims there were owls, diurnal birds of prey, charadriids and some passerines up to the size of a thrush.

In terms of the NISP and MNE, the most frequent bones of medium-sized grouse are the coracoid, humerus and scapula (Figure 1). The sternum and femur are half as numerous; the remaining bones are much rarer in the material. Remains of the skull and mandible are scarce. By far the best skeletal elements for the MNI calculation are coracoid, sternum, scapula and humerus—the MNI% values obtained on their basis range from 90% to 100% (Tables 2 and 3).

The elements of the axial skeleton are largely fragmented (Table 2). The sternum is usually damaged, but its front part is preserved with the rostrum sterni and the corpus sterni of different sizes, with jagged fracture edges (Figure 2a). In contrast, long bones are mostly complete; in most cases, the percentage of whole long bones exceeds 90%, (Table 3). The scapula, which is not a typical long bone, often survives as the epiphysis with a part of corpus scapulae of varying length.

3.2 Bone ratios based on the MNE

The wing bones were more numerous than the leg bones; proximal skeletal elements were more numerous than distal elements, whereas core elements were more numerous than limb elements (Table 4). The deviation from the expected proportions (6:6, 10:8 and 6:14, respectively) is statistically significant (p < 0.05) for each of the three ratios mentioned above (χ² = 29.776, χ² = 300.550 and χ² = 280.541, respectively). The differences between the data obtained on the basis of NISP and MNE were marginal (Table 4).

3.3 Perforations

Perforations attributable to the beak and/or claws were observed on approximately 10% of bones (Table 5). Individual skeletal elements were affected to a varying degree by punctures. The most frequently punctured element was the sternum (about 40% of all sterna were perforated), followed by the pelvis (about 25%), humerus and coracoid (about 10% each).

Almost half of the perforated humeri (9 out of 20) and every fifth sternum (9 out of 43) have two or more perforations (Figure 2a,b); multiple perforations were less frequently observed on other bones.

The coracoid was most often perforated in the sternal part (14 out of 16 perforated coracoids, Figure 2c); on the humerus, sternum and pelvis no higher frequency of perforation was found in any particular region.

A significant number of sterna were preserved in articulation with coracoids. Some of these combined elements show characteristic damage, which indicates that A. gentilis tends to hit specific body parts of its victims, in this case, the breast region (sternum-coracoid connection). This results in frequent damage to the processus craniolateralis of the sternum and the processus lateralis of the coracoid (Figure 2d,e). Altogether, about 77% of sterna (82 out of 106) have their processus craniolateralis damaged and about 48% coracoids (99 out of 207) have their processus lateralis damaged.

The share of immature specimens was about 8% (80 out of 981 bones).

4.1 Taxa composition

The composition of the fauna found in a given archaeological material can be a valuable clue about its origin (Bochenski, 2005; Lloveras, Cosso, Solé, Claramunt-López, & Nadal, 2018). What matters here is not only the species composition, which may vary depending on the geographic region and season, but also the size of the preferred prey, which is more constant for a given predator. In Fennoscandia, the main food of the goshawk is grouse, especially L. tetrix and L. lagopus, whereas corvids, thrushes, pigeons and hares and squirrels are important prey under certain circumstances (Tornberg et al., 2006). Thus, the food remains researched by us reflect well the actual goshawk diet in this region. The list of bird species in the northern goshawk's diet is very long, and today, grouse does not constitute its main food everywhere. Currently, the main food of the goshawk in Central Europe are partridges and doves (Brüll, 1977), but in historical times, grouse could also play a role similar to that in Fennoscandia.

4.2 Skeletal composition

The frequency of individual skeletal elements in archaeological materials may be an indication of its origin. In food remains deposited by man, usually, the humerus and femur dominate (Mourer-Chauviré, 1983). However, in food remains from birds of prey and owls, the situation is more complicated because it depends on several factors, including the size of the victim (Baales, 1992; Mourer-Chauviré, 1983) and whether the bones were swallowed and spit out as pellets or the meat was stripped of them and the bones themselves were not ingested (Bochenski, 2005). The coracoid and humerus predominate in the material currently studied, which matches the uneaten food remains of other large diurnal raptors including the imperial eagle, golden eagle, white-tailed eagle and gyrfalcon (Bochenski, 2005; Bochenski et al., 1999; Bochenski et al., 1997; Bochenski & Tornberg, 2003; Bramwell et al., 1987; Lloveras et al., 2018). These raptors, including the goshawk currently under study, also have a very low percentage of skulls and mandibles in uneaten food remains, and their prey's sterna, coracoids and humeri give very high MNI values. In addition to these species, also a few others, including the crested caracara and Egyptian vulture, have a large amount of sterna in their uneaten food remains (Lloveras, Nadal, et al., 2014; Montalvo et al., 2011).

4.3 Fragmentation

In terms of the degree of fragmentation of long bones, the studied food remains of the goshawk fit well with the Category 3 distinguished by Bochenski (2005), which includes non-ingested food remains of diurnal birds of prey. This category is characterized by a very low degree of fracture, with more than 60% of the bones intact. Most species of diurnal birds of prey are included in it, which is confirmed by various independent studies (Bochenski et al., 2009; Bochenski & Tornberg, 2003; Laroulandie, 2002; Lloveras et al., 2018; Lloveras, Nadal, et al., 2014). The large fragments of the sternum with the preserved rostrum sterni and the jagged edges of the corpus sterni are also typical of the non-ingested food remains of various diurnal birds of prey (Bochenski et al., 2009; Lloveras et al., 2018; Lloveras, Nadal, et al., 2014), gulls (Serjeantson, Irving, & Hamilton-Dyer, 1993) and possibly other avian predators.

4.4 Ratios

The results obtained from all three bone ratios (wing/leg, proximal/distal and core/limb) fully agree with the bone damage found in non-ingested food remains of diurnal birds of prey (Bochenski, 2005). The wing/leg ratio: in most species of diurnal birds of prey, uneaten food remains are dominated by wing bones (Bochenski et al., 1997; Laroulandie, 2002; Lloveras, Nadal, et al., 2014; Lloveras, Thomas, et al., 2014; Montalvo et al., 2011) although there are also exceptions, such as for example, the Golden Eagle, where a slight predominance of leg bones was observed (Lloveras et al., 2018). In pellet materials, the wing/leg ratio is usually close to 1:1 (Bochenski, 2005) whereas assemblages accumulated by man can have very different wing/leg ratios, depending on for instance the purpose for which the birds were collected, hunting method, differential transport of carcases or the properties of the bones themselves (Bovy, 2002, 2012; Ericson, 1987; Laroulandie, 2010; Lefèvre & Laroulandie, 2014). The proximal/distal element ratio: high preponderance of proximal elements in food remains of the northern goshawk corresponds to Category 3 predators, which include the golden eagle (Bochenski, 2005). It is noteworthy that a similarly high dominance of proximal elements was also found independently in studies on food remains of the golden eagle (Lloveras et al., 2018). Pellets of diurnal birds of prey and owls show either an equal share of proximal and distal elements (Category 1 predators) or a slight dominance of proximal elements (Category 2 predators), which in turn allows to distinguish between these three groups of predators (Bochenski, 2005). The core/limb ratio distinguishes between pellet materials of owls and diurnal raptors (limb elements prevail) and non-ingested food remains of diurnal raptors (core elements prevail) (Bochenski, 2005). The currently studied northern goshawk material fits this pattern very well, because core elements prevail to a similar degree as in several independent studies on golden eagles (Bochenski et al., 1999; Bramwell et al., 1987; Lloveras et al., 2018).

4.5 Perforations

The perforations found on the bones of the northern goshawk victims are similar to those left by other species of raptors. Possible differences may be in the percentage of damaged bones. Current results indicate that the northern goshawk perforates the bones of its victims to a rather moderate degree (about 10%). A similarly low percentage of perforated bones was found in Bonelli's eagle victims (approximately 6%; Lloveras, Thomas, et al., 2014), slightly higher in golden eagle victims (approximately 14%; Lloveras et al., 2018) and the highest in Egyptian vulture (approximately 28%; Lloveras, Nadal, et al., 2014). A common feature of all species of birds of prey that have so far been examined for bone perforation of their victims is the high frequency of perforations on the sternum, pelvis, humerus and coracoid (Bochenski et al., 2009; Laroulandie, 2000, 2002; Lloveras et al., 2018). The northern goshawk, currently under study, is no exception. For two reasons, it can be expected that the sternum will be most prone to perforation. First, there is a lot of meat on the sternum, and second, it is an element of the skeleton with a relatively large surface area. Bochenski et al. (2009) have already suggested that the shape and surface of the prey sternum may play a significant role in the amount of perforations. They found that the wide sternum of ducks was more often perforated than the narrow sternum of galliforms.

The simultaneous damage to the sternum by breaking the processus craniolaterales and the coracoid by breaking the processus lateralis was noticed due to the fact that many food remains in the studied material, including the sternum-coracoid complex, were still in articulation. Such damage has also been found on many isolated sterna and corcaoids. This damage has not been described anywhere before. It seems that it may also be characteristic of other species of birds of prey. Due to the high frequency of sterna and coracoids damaged in this way, it can be expected that they will also be visible on bones from archaeological excavations.

Current studies, including skeletal composition, fragmentation and perforations, indicate that the goshawk damages the bones of its prey in a manner typical of diurnal raptors. Bird bones from open-air archaeological sites that were temporarily or alternately occupied by man may contain food remains of the goshawk, whose habitat is forests in the vicinity of open areas. This study will assist in determining the factors responsible for the accumulation of zooarchaeological materials.