Predation on eggs and nestlings is a primary cause of reproductive failure among birds, thus many species exhibit defense behavior when predators approach the nest. Nest defense behavior may reflect an optimization of costs and benefits to the parent birds' fitness in relation to current versus future reproduction (Barash 1975, Montgomerie and Weatherhead 1988), yet large variation in the extent of nest defense exists among and within species. Factors that may influence the type and degree of nest defense include nest type and nesting habitat or site (Curio et al. 1985, Albrecht and Klvana 2004), food abundance (Tolonen and Korpimäki 1995), predator type including humans (Brunton 1990, Winkler 1992), offspring value (Grieg-Smith 1980, Olendorf and Robinson 2000), and risk factors to either the young or defending parent (Regelmann and Curio 1983, Krüger 2002).

Nest defense by birds of prey may differ from that of other birds because raptors can potentially attack and injure would-be predators. In addition, raptor young can be considered to have relatively high value because many species reproduce, on average, once annually or less and have relatively few young, and for some species, a scarcity of safe nest sites may limit breeding density and success (Newton 1979, Village 1983). Many raptors are also highly sensitive to human disturbance, possibly because of continued persecution in some areas (Newton 1979). Thus, as a group, raptors may be expected to exhibit relatively aggressive nest defense.

Numerous studies of avian nest defense have focused on single species and factors potentially influencing the type and extent of defense behaviors (reviewed in Redondo 1989). By seeking repeated patterns throughout a broad taxonomic group, comparative studies of behavior can lend insight into evolutionary and ecological factors potentially underlying interspecific variation (for examples see Röell and Bossema 1982, Larsen et al. 1996, Meilvang et al. 1997, Gunness and Weatherhead 2002). We used a comparative approach (Harvey and Pagel 1991) to explore patterns of nest defense exhibited by diurnal North American raptors and to assess factors that might influence these behaviors. Our underlying hypotheses were: (1) parental defense decisions are influenced by risk to the developing young or risk to the defending parent, and (2) more aggressive defense would be exhibited by species for which young are more vulnerable or of greater value.

Methods

We defined a raptor's nesting period as the time during which adults are tending an occupied nest (includes egg-laying, incubation, and nestling stages). We identified four categories of behavior (sensu Hudson and Newborn 1990) potentially exhibited by a raptor when a predator approaches the nest: (1) no defense, bird may fly away (no defense), (2) circles or calls when predator approaches (passive response), (3) dives at or chases predator, but makes no physical contact (somewhat aggressive), and (4) physically attacks predator (very aggressive). Using information from journal articles, review papers, and books, we recorded accounts of nest defense for diurnal raptors in North America and assigned each account to one of the above four categories based on behavioral descriptions. We defined “account” as any mention of nest defense behavior by the author(s). If more than one defense behavior was noted for the same individual during the same observation, we recorded it as one account and classified it according to the most aggressive behavior noted. If the author(s) reported defense behavior exhibited by another individual or by the same individual on another day or against a different predator, we recorded those observations as separate accounts of nest defense. For example, if a raptor called, chased, or attacked a predator during three separate observations, we gave that species one account of nest defense for call/circle, one account for dive/chase, and one account for attack.

After compiling all accounts for each raptor species in the dataset, we included only those species having four or more accounts of nest defense in subsequent analyses. We assigned each species to one overall defense category (no defense, passive response, somewhat aggressive, or very aggressive) based on the category with the highest frequency of accounts for that species recorded from all data sources; we assumed this category represented the “typical” defense behavior of that species. If more than one category had the same number of accounts, we assigned the species to the more aggressive category.

A priori, we identified four factors that may influence nest defense of raptors. First, we examined reproductive effort, predicting that raptors with high reproductive effort throughout the nesting period defend nests more aggressively than species with low reproductive effort throughout the nesting period (Redondo 1989). Second, we examined the cover types where nesting occurred, predicting that raptors nesting in open cover types exhibit less aggressive or no defense because such behaviors might be overtly conspicuous to visually-oriented predators leading to high rates of nest discovery and loss (Carillo and Aparicio 2001, Bures and Pavel 2003). Third, we examined nest accessibility, predicting that nests easily accessed by predators (ground and tree nests) are defended more aggressively than nests that are more difficult to access (cliff and cavity nests), because accessible nests incur higher predation rates (e.g., Martin 1995, Wesolowski and Tomialojć 2005). Finally, we examined responses by raptor species to different predator types, predicting that raptors exhibit less aggressive nest defense toward humans than toward other mammalian or avian predators because raptors may recognize humans either as non-predators or as large predators against which the response risk to themselves is high (Knight et al. 1989, Galeotti et al. 2000).

We developed two indices of reproductive effort. One represented effort during the early part of the nesting period (egg-laying stage) and one represented effort in the latter part (incubation and nestling stages). Early reproductive effort was the investment required to produce a clutch calculated as average egg volume times average clutch size. We calculated egg volume as K_v × LB², where L  =  length and B  =  breadth of the egg, respectively (in mm), and K_v  =  0.0005 (for oval eggs; Hoyt 1979). We defined reproductive effort in the latter part of the nesting period as time spent incubating eggs and raising nestlings. We calculated this index as the sum of the mean incubation period and the mean nestling period (both in days) for each species. Because both measures were correlated with female body mass, we used the residuals of the regression analyses of early and late reproductive effort, respectively, on log female body mass as the indices in subsequent analyses. We calculated both indices for each species and then ranked each index separately using a ranking function in Microsoft Excel© (Microsoft Corp. 2000, Redmond, WA U.S.A). We classified species for which the residuals were ≤0 as having low reproductive effort and species for which the residuals were >0 as having high reproductive effort, for both the early and latter part of the nesting period. We obtained information on clutch size, length of incubation and nestling periods, and female body mass from species accounts in the Birds of North America series (Poole and Gill 2002).

We classified nesting cover type for each species as either open (including grasslands, tundra, deserts, and areas with scattered trees or shelterbelts) or closed (including all forested communities containing deciduous, coniferous, or mixed tree species with a predominately closed canopy cover). We defined nests for each species as either accessible (ground and tree nests) or inaccessible (cliff and cavity nests). For early and late reproductive effort, nesting cover type, and nest accessibility, we examined proportions of raptor species in our dataset that have either no defense, passive response, somewhat aggressive, or very aggressive nest defense.

We examined responses by raptors to a suite of potential predators. For each account of defensive behavior, when possible, we identified the predator against which the behavior was directed as (1) diurnal avian (e.g., crows [Corvus spp.]), (2) nocturnal avian (e.g., owls), (3) mammalian (not human, e.g., squirrels), and (4) human. Then we classified each account for which the predator could be identified into one of the aforementioned four nest defense categories.

Although our initial investigations were conducted at the species level, we recognized that evolution of behavior may have a strong historical component and that data sampled across several closely-related species may not be independent (Harvey and Pagel 1991, Freckleton 2000). Therefore, to help control for phylogenetic constraints, we further investigated patterns in nest defense at the genus level, for early and late reproductive effort, nesting cover type and nest accessibility, for all raptors in our dataset. Because of uncertainty in phylogenetic relationships within Falconiformes (e.g., Griffiths et al. 2004, Kruckenhauser et al. 2004), we also conducted a qualitative examination of possible phylogenetic constraints in our evaluation of raptor nest defense by mapping these behavior patterns on available phylogenies for birds of prey (Griffiths 1999, Riesing et al. 2003).

Results

There were 19 species of diurnal North American raptors for which we found at least four accounts of nest defense (Table 1). For some, including Crested Caracara (Caracara cheriway), Bald Eagle (Haliaeetus leucocephalus), Golden Eagle (Aquila chrysaetos), and White-tailed Hawk (Buteo albicaudatus), we recorded accounts in which the raptor simply left the nest area upon approach of a predator. However, we classified these raptors as having “passive response” nest defense because there were accounts describing calling and circling behavior against a predator for all four species (Table 1). We classified all other species as having either somewhat aggressive or very aggressive nest defense (Table 1).

Table 1

Information on factors potentially influencing nest defense of diurnal North American raptors for which we found at least four accounts of nest defense in the literature (N  =  19 species). Categories of nest defense are: (1) no defense, bird may fly away (no defense), (2) circles or calls when predator approaches (passive response), (3) dives at or chases predator, but makes no physical contact (somewhat aggressive) and (4) physically attacks predator (very aggressive). Reproductive effort: H  =  high, L  =  low. Nesting cover type: O  =  open, C  =  closed (see text for further descriptions). Nest accessibility: A  =  accessible (ground or tree), I  =  inaccessible (cliff or cavity).

Phylogeny

When we mapped nest defense for raptors in our dataset on available phylogenies (Griffiths 1999, Riesing et al. 2003), we found that except for Merlins (Fig. 1a), raptors in the genus Falco exhibit very aggressive nest defense, and all have inaccessible nests. Another exception in Falconidae was the relatively large-bodied Crested Caracara (Fig. 1a), a tree nester in the subfamily Caracarinae (Griffiths 1999) that exhibits little nest defense. Northern Goshawks (Accipiter gentilis) and Cooper's Hawks (in the outgroup Accipiter, Griffiths 1999), which have open, accessible nests and nest in closed cover types exhibit very aggressive nest defense. These two species also have high reproductive effort early in the nesting period. The relatively smaller-bodied Red-shouldered Hawk (Buteo lineatus) had very aggressive nest defense, and the White-tailed Hawk had passive response nest defense, while other Buteos in our dataset had somewhat aggressive nest defense (Fig. 1b). Red-shouldered Hawks are one of only two Buteo species that have high reproductive effort throughout the nesting period, whereas the relatively larger-bodied White-tailed Hawk has low reproductive effort early in the nesting period (Fig. 1b, Table 1).

Figure 1

Nest defense behavior, nesting cover type, and nest type of 15 diurnal raptors mapped onto schematic phylogenies adapted from (a) Griffiths (1999) and (b) Riesing et al. (2003).

Phylogenetic information was not available for Osprey, Bald Eagle, Gyrfalcon, and Northern Harrier. Nest defense behavior:  =  very aggressive,  =  somewhat aggressive, ░  =  passive response. Nesting cover type: •  =  closed cover type,  =  open cover type. Nest type:  =  cliffs or cavities (inaccessible),  =  tree or ground nest (accessible). Line lengths are not to scale and dashed lines indicate levels or connections that are not shown.

Predator Type

Accounts of raptor nest defense for which the potential predator could be identified suggest differences in response to different predator types (Fig. 2). There were more accounts of somewhat aggressive or very aggressive defense behaviors (dive/chase and attack) against both diurnal and nocturnal avian predators (N  =  61 accounts and N  =  10 accounts, respectively), but more accounts noting less aggressive behaviors (call/circle) or no defense against potential human predators (N  =  37 accounts). There were similar proportions of accounts in each defense category exhibited by raptors when defending against non-human mammalian predators, although there were only eight of these accounts.

Figure 2

Responses by raptors to different predator types, N  =  19 species, 116 accounts.

Numbers above the bars indicate number of accounts within each nest defense category.

Discussion

We attempted to describe patterns in nest defense for a variety of diurnal North American raptors and to identify factors that may influence the types and expression of these behaviors. Our results suggest, as for many other birds, that a variety of factors affect nest defense of raptors, yet assessment of their relative influence is likely confounded by interactions among them. For example, any influence of reproductive effort on nest defense is likely complicated by body size. Relationships between body size and antipredator strategies are well documented among birds (Andersson and Norberg 1981, Wiklund and Stigh 1983) and across taxonomic groups (Larsen et al. 1996); typically larger species exhibit more aggressive nest defense. In our study, species that exhibit the highest levels of nest defense (those in the genus Falco, the two accipiters, and the Red-shouldered Hawk), are small-bodied relative to other species in our sample and are the species best adapted for fast, highly maneuverable flight. These characteristics may afford them reduced risk of injury from a potential nest predator, suggesting the hypothesis that nest defense is influenced by flying ability.

Overall, the larger raptors showed less aggressive nest defense. They are not fast flyers and their size may deter predators before an attack occurs; thus, aggressive nest defense may not be as necessary to deter predators (Andersson and Norberg 1981, Wiklund and Stigh 1983). However, in contrast to this pattern, the Ferruginous Hawk (Buteo regalis) exhibits somewhat aggressive nest defense. If young of this species have particularly high reproductive value (high reproductive effort throughout the nesting period; Table 1), and the probability of nest loss to a predator is high (this hawk typically nests in open cover types on or close to the ground), parents are expected to show more aggressive nest defense (i.e., the Reproductive Value-Stimulus Value hypothesis; Patterson et al. 1980) because potential risk to the parents may be offset by increased offspring security (Andersson et al. 1980). Similarly, Swainson's Hawks (Buteo swainsoni) nest in sparse shrubs or trees occurring in open cover types, nest sites that may be limiting as they are frequently subject to takeover by other species such as Red-tailed Hawks (Buteo jamaicensis), Common Ravens (Corvus corax), and crows (Cottrell 1982, England et al. 1997). Competition for limited, safe nest sites may lead to more aggressive nest defense by adult Swainson's Hawks.

Interactions between nest type and nesting habitat probably also influence nest defense. We originally predicted that raptors nesting in open cover types would not exhibit aggressive nest defense because such behaviors may be overtly conspicuous to visually-oriented predators leading to high rates of nest discovery and loss (Carillo and Aparicio 2001, Bures and Pavel 2003). Our results did not support this prediction overall, but exceptions were noted. For example, in open grasslands in Florida, Crested Caracaras show little nest defense (J. Morrison unpubl. data). Crows, common nest predators in that landscape operate in groups; thus, efforts by a pair of caracaras to defend their nest may not be worth the potential risk to themselves from these aggressive social predators.

The raptors in our dataset differed in their responses to different predator types, which is similar to findings for other avian species. Typically, the intensity of nest defense and tendencies toward risk vary by predator type and length of time the birds had been exposed to the predator (Knight 1984, Gottfried et al. 1985, Brunton 1990, Ferrer et al. 1990, Halupka 1999). Our finding that raptors exhibit less aggressive nest defense against potential human predators supports our original prediction and results of other studies. Knight et al. (1989) found that call and dive rates for Red-tailed Hawks were highest in areas more recently settled by humans and lowest in sites settled the longest, suggesting habituation to humans. Long-eared Owls (Asio otus) experiencing higher levels of human persecution show less aggressive nest defense than owls breeding in undisturbed areas perhaps because the former have become unwilling to take risks against human predators (Galeotti et al. 2000). Nesting birds worldwide probably perceive humans as a serious threat, but it is likely that this perception varies greatly with differences in human behavior. Where birds experience low levels of threat from humans, parents may exhibit high levels of nest defense (Knight et al. 1987, Ferrer et al. 1990). By contrast, where nesting birds are frequently shot or trapped, parental defense may be too costly. Thus, low levels of nest defense against humans may be the most frequent strategy in areas of intense human pressures.