Portal de Periódicos da CAPES

Attacks of songbirds in mixed‐species flocks by Eurasian Sparrowhawks: strategies of predators and potential prey

2020; Association of Field Ornithologists; Volume: 91; Issue: 4 Linguagem: Inglês

10.1111/jofo.12350

1557-9263

Indriķis Krams, Tatjana Krama, Todd M. Freeberg, Ronalds Krams, Kathryn E. Sieving,

Rangeland and Wildlife Management

Predation is an essential factor affecting prey populations, yet attacks are notoriously difficult to observe in real time. Here, we provide descriptive data about the attack behavior of Eurasian Sparrowhawks (Accipiter nisus) and the escape tactics of their prey in coniferous forests in Latvia during the non-breeding season. Over a period of 36 years (1985–2020), we observed 199 attacks of Eurasian Sparrowhawks on mixed-species flocks of small forest passerines, with 19 attacks being successful (9.6%). Sparrowhawks attacked mainly from either above the canopy of mature conifers or just below the lower branches of the canopy. Attacks by sparrowhawks were more successful in the lower canopy (N = 17) than the upper canopy (N = 2). Fourteen of the 19 songbirds killed by sparrowhawks were Goldcrests (Regulus regulus), Willow Tits (Poecile montanus), and Coal Tits (Periparus ater), small songbirds typically limited to the lower canopy by larger, more dominant species that occupy the upper canopy. Sparrowhawk attacks on birds located in the upper canopy were rarely successful, likely because songbirds in the upper canopy are more concealed by the vegetation and more likely to detect approaching predators. Our results suggest that the spatial segregation of members of mixed-species groups of wintering songbirds can be explained by their responses to microhabitat (canopy height) gradients in predator exposure that are enforced by group dominance hierarchies. Ataques de aves passeriformes en bandadas de especies mixtas por parte del gavilan euroasiático (Accipiter nisus): estrategias de los depredadores y presas potenciales La depredación es un factor esencial que afecta a las poblaciones de las presas, pero los ataques son notoriamente difíciles de observar en tiempo real. Aquí presentamos datos descriptivos sobre el comportamiento de ataque de los gavilanes euroasiáticos (Accipiter nisus) y las tácticas de escape de sus presas en los bosques de coníferas de Letonia durante la época no reproductiva. En un período de 36 años (1985–2020), observamos 199 ataques de gavilanes euroasiáticos a bandadas de especies mixtas de pequeños paseriformes de bosque, 19 ataques tuvieron éxito (9,6%). Los gavilanes atacaron principalmente desde arriba del dosel de las coníferas maduras o desde justo debajo de las ramas inferiores del dosel. Los ataques de los gavilanes tuvieron más éxito en el dosel inferior (N = 17) que en el superior (N = 2). Catorce de los 19 individuos de aves paseriformes muertas por los gavilanes pertenecieron a Regulus regulus, Poecile montanus y Periparus ater, las aves de tamaños pequeños comúnmente están limitadas al dosel inferior por especies más grandes y dominantes que ocupan el dosel superior. Los ataques del gavilán a las aves situadas en el dosel superior rara vez tuvieron éxito, probablemente porque las aves paseriformes del dosel superior están más ocultas por la vegetación y es más probable que detecten a los depredadores que se acercan. Nuestros resultados sugieren que la segregación espacial de los miembros de grupos mixtos de aves cantoras durante el invierno puede explicarse por sus respuestas a los gradientes de microhábitat (altura del dosel) en la exposición a los depredadores que está reforzada por las jerarquías de dominancia del grupo. Survival is an essential component of life history that affects fitness (Clutton-Brock 1988, Moiron et al. 2020), whereas predation is a selective force that exerts influence on the fitness of prey individuals and their community structure and population dynamics (Kullberg and Ekman 2003, Goodale et al. 2019, Meise et al. 2020). Predation acts through either consumption of prey individuals or non-consumptive effects influencing their metabolism, behavior, physiological mechanisms of adaptations, and morphological phenotypes (Thomson et al. 2010, Krams et al. 2013a,b, Turbill et al. 2019, Yin et al. 2019, Zanette et al. 2019). Most recent research has focused on the non-consumptive effects of predation because the results of a number of studies have shown that the non-consumptive effect of predators on prey behavior and community and population structure and dynamics can equal the effects of direct consumption (Preisser et al. 2005, Abrams 2007, Creel and Christianson 2008, Van Dievel et al. 2019). However, the observed focus on non-consumptive effects can also be explained by the scarcity of observations of predation events because predators are notoriously difficult to study. Although predator attacks and actual killings are difficult-to-witness behaviors (Quinn and Cresswell 2004, Roth et al. 2008), these data are needed to assess how prey vulnerability mediates natural selection in prey populations. Although there are many studies of ecology and behavior of wintering parids, few investigators have studied the hunting strategies of their predators. Members of mixed-species groups of birds are known to respond to variation in perceived predation risk by changing their foraging niches in trees (Ekman 1986, Krams 1996, Roth and Lima 2007a, b). For example, the risk of predation by Pygmy Owls (Glaucidium passerinum) varies within the tree canopy, being higher for birds that use the outer parts of branches and the lower parts of mature trees (Ekman 1986, Kullberg 1995). Accordingly, socially dominant individuals in winter groups of birds tend to occupy the safest canopy parts and subordinates are expelled to microhabitats where they are more vulnerable to predators (Kullberg 1998, Kullberg and Ekman 2003). During winter, Eurasian Sparrowhawks (Accipiter nisus) are keystone predators on small birds of forests in Europe (Morse 1973, Perrins 1979, Götmark and Post 1996, Krams 1996, Zawadzka and Zawadzki 2001) and are known to affect prey behavior and foraging site selection (Carlson et al. 2017, Kalb et al. 2019). Eurasian Sparrowhawks rely on surprise, using cover and irregularities in terrain to secure a close approach (Newton 1986, Hinsley et al. 1995). However, little is known about how a bird’s location in different parts of the forest canopy might affect their risk of predation by attacking Eurasian Sparrowhawks (Krams 1996, 2001a) or how the responses of birds attacked in different parts of the canopy might differ. To address these gaps in our understanding of predator–prey interactions, we provide descriptive data on the attacking and killing behavior of Eurasian Sparrowhawks and the escape tactics of their prey during the non-breeding season. Observational data were collected during 35 consecutive seasons from September to early March (1985–2020) near the town of Krāslava (55°87'N, 27°19'E) in southeastern Latvia (Rytkönen and Krams 2003). Attacks on members of mixed-species flocks of birds by Eurasian Sparrowhawks (hereafter, sparrowhawks) were observed in 60- to 110-year-old coniferous forests dominated by Scots pine (Pinus sylvestris), with admixture of Norway spruce (Picea abies) and silver birch (Betula pendula). Observations were made primarily during field studies of wintering parids (Krams 1996, 1998a,b, 2000a,b, Krams et al. 2001, 2013c, Krama et al. 2008, 2015, Cīrule et al. 2017). We defined a mixed-species group of parids as “a group of independently moving birds of the same trophic level from more than one species found in close proximity, which interact with one another” (Goodale et al. 2019). In our study area, mixed-species groups were typically composed of Goldcrests (Regulus regulus; N [the typical number of individuals in a mixed-species group] = 4–6, body mass = 5.5 g), Marsh Tits (Poecile palustris; N = 3–4, body mass = 12 g), Willow Tits (Poecile montanus; N = 4–5, body mass = 11 g), Crested Tits (Lophophanes cristatus; N = 4–5, body mass = 11 g), Coal Tits (Periparus ater; N = 3–6, body mass = 9.2 g), Blue Tits (Cyanistes caeruleus; N = 2–4, body mass = 10.5 g), Great Tits (Parus major; N = 1–6, body mass = 18.2 g), Eurasian Nuthatches (Sitta europaea; N = 2–3, body mass = 20.5 g), Eurasian Treecreepers (Certhia familiaris; N = 1–3, body mass = 10 g), and, occasionally, Chaffinches (Fringilla coelebs; N = 3, body mass = 21 g). For each attack by a sparrowhawk, we estimated the foraging height of the bird being attacked, tree height, and canopy height where the attacked bird was either captured or escaped. Foraging heights are normally expressed as proportions of canopy height and divided into four categories (Ekman 1986, Krams 1996, Cīrule et al. 2017). In many cases, however, we were unable to determine the location in the canopy where predators killed their prey or where prey escaped. Importantly, most observed attacks surprised not only the prey, but the observers because sparrowhawks approach prey at high speed (Barnard 1979). Therefore, relative foraging height was categorized as either the upper part or lower part of the canopy. We identified the species of each attacked individual bird and distinguished between successful and unsuccessful attacks. Time of the day was also recorded. For each attack, we also identified the sex of a sparrowhawk. Males and females are easy to tell apart (Newton 1986). Females have brown backs and wings with brown stripes on the chest and belly, whereas males have blue-gray backs and wings with orange stripes on the chest and belly. Female sparrowhawks are up to 25% larger than males. We observed 199 attacks by sparrowhawks (~0.02 attacks per flock/h during 12,500 h of observations). Of these attacks, sparrowhawks killed their prey in 19 cases (9.6%) and failed 180 times (90.4%). All attacks were by male sparrowhawks, and all attacks were in the canopy of pines. More attacks were observed in the morning (between dawn and 11:00) (N = 88 and 10 killed prey), fewer during the middle of the day (between 11:00 and 14:00) (N = 36 and two killed prey), and more again at the end of the day (between 14:00 and sunset) (N = 75 and seven killed prey). Goldcrests and Coal Tits were attacked and killed most often, followed by Willow Tits (Table 1). For five other species, only one individual was killed by sparrowhawks, including Marsh Tits, Crested Tits, Blue Tits, Great Tits, and Chaffinches. Relative attack rates (attacks per number of individuals of each species in one mixed-species group) also show that Goldcrests, Coal Tits, and Willow Tits were the most vulnerable species in the mixed-species groups (Table 1). We observed no attacks of either Eurasian Nuthatches or Eurasian Treecreepers. When attacked in the upper part of the canopy, birds were observed to dive down to the lower canopy and then ascend to escape in the canopies of adjacent trees (65 of 75 observations). Sparrowhawks usually gave up as soon as the birds being attacked started escaping upward. One Crested Tit tried to escape by horizontal flight and was killed, and one Willow Tit was captured while excavating food from a tree trunk. When attacked in the lower part of the canopy, birds often dived down to ground level to try to escape in bushes and small pines or spruces (35 of 87 observations) where sparrowhawks ceased their attacks. In 43 cases, birds escaped either by making upward flights at a steep angle to the upper parts of the canopy of the same tree or by making horizontal flights to escape in the upper parts of the canopy of a neighboring tree (9 of 87 cases). With the exception of Eurasian Nuthatches, species in our study had similar escape strategies. Eurasian Nuthatches foraging on tree trunks avoided predation by moving to the opposite side of trunks (N = 3), and one nuthatch hid in the dense ground cover of small spruces. After attacks, birds remained motionless for an average of 4.63 ± 1.95 (SD) min. Sparrowhawks always relied on surprise and approached prey from either below the canopy or just above the uppermost parts of the canopy (Table 1). Overall, sparrowhawks initiated 84 attacks in the upper part and 115 attacks in the lower part of the canopy. However, sparrowhawks killed significantly fewer prey (N = 2) in the upper canopy than in the lower canopy (N = 17; Fisher exact test, P = 0.0029). Goldcrests (0 vs. 4) and Coal Tits (0 vs. 6) were also killed more often in the lower canopy (Fisher exact tests, P = 0.049 and P = 0.035, respectively) (Table 1). The other species did not differ in the number of individuals killed in the upper and lower parts of the canopy (Fisher exact tests, all P > 0.05). We found that small passerines were the main winter prey of sparrowhawks in our study area. Similar results have been reported at other locations in Europe (Cramp and Simmons 1980, Newton 1986, Götmark and Post 1996, Rytkönen et al. 1998). Hunting methods mentioned in the literature involve surprise attacks of small birds, usually located in woodland edges, glades, clearings, and other continuous or adjacent lines and clusters of cover in partly open terrain (Newton 1986). Accipiters in urban areas frequently use buildings as cover when making surprise attacks of birds at feeding stations (Malone et al. 2017, K. E. Sieving, pers. observ.). However, our study site offered few such opportunities to hide approaches, and sparrowhawks attacked mainly by flying either above the canopy of mature conifers or just below the lower branches. Small birds in the upper part of the canopy would likely be concealed by the vegetation because the closed structure of branches provides shelter from airborne predators (Krams 1996). This may explain the low success rate of sparrowhawks in our study, especially for attacks in the upper canopy. Although the total number of attacks was low, the capture rate of sparrowhawks was within the range observed in single-species systems outside the breeding season (Cresswell 1994, Cresswell et al. 2002) or lower than the capture rate observed in mixed-species groups (Roth and Lima 2003). The mixed-species groups observed by Roth and Lima (2003) were short-term social associations with an open membership influenced by predation and formed around a common resource (Goodale et al. 2019, 2020). The mixed-species groups attacked by sparrowhawks in our study were more coherent, permanent, nearly year-round associations with a closed membership. Indeed, our data support the general consensus that individuals participating in highly organized mixed-species groups accrue effective protection and, in turn, this explains the low capture rate of sparrowhawks. Under the assumption that dominant flock members will occupy the safest foraging sites in trees, the patterns we observed can be interpreted as follows. In the lower canopy, sparrowhawks attacked and killed mostly Goldcrests, Willow Tits, and Coal Tits. Although these three species are the most abundant in the mixed-species groups, other species such as Crested Tits and Great Tits are often equally numerous (Krams et al. 2020). Therefore, differences in the abundance of species alone cannot explain the higher mortality of Goldcrests. Given that food resources in general, and arthropods in particular, are most abundant in the higher canopy (Suhonen et al. 1992, Krams et al. 2001) that is normally occupied by the larger and more dominant species in the flocks, the species attacked most often are likely forced into a greater relative vulnerability in the lower canopy. Body size determines social dominance relationships in these groups (Krams 1998a), and the smaller Goldcrests (the smallest European bird), Willow Tits, and Coal Tits are subordinate to Crested Tits, Great Tits, and Eurasian Nuthatches (Krams et al. 2020). Foraging site selection in mixed-species groups and the parts of trees used by individual birds are influenced by a combination of the dominance relationships within flocks, food availability, and predation risk (Pulliam and Caraco 1984). In general, smaller species are expelled by dominant species from the more protected and food-rich microhabitats (Hogstad 1988a,b, Krams 1996) in mixed-species groups with closed memberships. Thus, foraging in the lower parts of the canopy may in part explain the prevalence of smaller passerines in the diet of male sparrowhawks. However, sparrowhawks in our study often attacked prey in the upper canopy even though kill rates were higher in the lower canopy. Male sparrowhawks may initiate attacks in the upper canopy, despite poor success, because the dominant species are also the largest species, and Accipiters prefer to prey on larger-bodied birds (Malone et al. 2017). Several investigators have found that the upper canopy is the preferred foraging site of dominant members of mixed-species tit groups because of increased availability of food and increased safety (Ekman and Askenmo 1984, Ekman 1987, Krams et al. 2001). In part, the mechanics of collective detection in these flocks likely influences the observed survival asymmetry associated with vertical foraging locations (Ekman and Askenmo 1984, Koivula et al. 1996). The arboreal environment of birds in mixed-species groups is unique from the viewpoint of predator detection, with the refuge and foraging substrate unavoidably obstructing the visual detection of predators. Therefore, a sparrowhawk attacking members of the mixed-species groups in the upper canopy would likely be first detected by dominant individuals foraging near treetops, assuming that dominant individuals maintain high levels of vigilance (Lima and Zollner 1996, Krams 1998b, Lima and Bednekoff 1999). Mixed-species flocks consist of species that differ in visual and auditory acuity (Lucas et al. 2007, Nolen and Lucas 2009, Jones and Sieving 2019). Thus, a diversity of species in these flocks may provide greater collective ability to detect predators, with the result that both dominant and subordinate species benefit from each other’s presence (Heymann and Buchanan-Smith 1990). Our observations suggest that dominant individuals remain in the upper half of canopies because of a “detection effect,” with vigilant birds usually reacting faster and avoiding predation by flying to cover (Cresswell 1994, Lima 1994). Birds foraging near the top of trees are at less risk of predation if an attack is detected at its start, not in the final phase. Early detection from a conspicuous foraging site such as the upper canopy provides the detector, often a dominant individual, with sufficient time to take cover (Krams 2001b). Dominant flock members often rely on their own vigilance in addition to the collective vigilance of flock members (Krams 2001a). Therefore, foraging near treetops might be the best strategy for improving survival because sparrowhawks would not be able to implement their most successful hunting tactic based on surprise. Finally, dominant individuals in the upper canopy not only enjoy better foraging conditions and increased safety based on both collective and personal vigilance (Lima 1994, Krams 2001a), but also efficiency in vocal communication and this, in turn, can provide benefits for all other flock members. Dominant male Crested Tits give loud and audibly conspicuous trills near the treetops, likely because this location increases the distance that sound can travel (Marten and Marler 1977). Moreover, warning calls from vigilant dominants at the treetops would transmit well, helping subordinate species foraging lower in the canopy to escape. The anti-predator signaling system of parids is finely tuned, encoding detailed situationally specific information about threats (Jones and Sieving 2019). For example, parids vary their vocal behavior in response to varying levels of risk associated with different predators (Krams et al. 2012), including flight and head orientation of predators (Zachau and Freeberg 2012, Kyle and Freeberg 2016), as well as different predator locations during attack (upper versus lower canopy; Freeberg 2008, unpubl. data). In sum, because predation by sparrowhawks is a force shaping use of foraging niches by members of mixed-species groups with a closed social system (Ekman 1989, Forsman et al. 1998), avoiding predation by sparrowhawks while foraging in one of the most exposed parts of the tree canopy is important, not only for the dominants who survive better, but for the vigilance and warning calls provided by dominants to subordinates in the more vulnerable sites. Different species in mixed-species flocks differ in their likelihood of survival, but the vulnerability of any of the species outside of the shared vigilance afforded by these close-knit social groups is likely intolerably higher (Goodale et al. 2020). This work was supported by a grant (lzp-2018/2-00057) from the Latvian Council of Science.