Predation is the main cause of nest failure for many bird species (Martin 1993), and nest survival is an important component of fitness (Lack 1966; Saether and Bakke 2000). Consequently, predation of nests has shaped the evolution of avian behaviors such as nest-site selection and parental attendance (Ghalambor and Martin 2002; Peluc et al. 2008), life history characteristics such as clutch size (Martin 1995), and morphological traits such as egg color (Kilner 2006). Nest predation also shapes population growth (Saether and Bakke 2000) and community structure by favoring nest-site diversification to reduce competition for predator-free space (Lima and Valone 1991). Therefore, ornithologists study nest predation to better understand the evolution and ecology of birds.
An understanding of how and why nest predation occurs requires examination of the predation process (Lahti 2009). Nest predation involves interaction between predator and prey, so ecological traits of predators, namely their abundance and behavior, determine predation risk (Thompson 2007). Accordingly, several studies link predator ecology with predation rates and patterns (Schmidt and Ostfeld 2003a,b; Sperry et al. 2008; Weatherhead et al. 2010). Nesting parent birds also influence predation risk by deciding where to nest (Martin 1998; Davis 2005; Peluc et al. 2008; Latif et al. 2012), modulating activity at the nest and consequently the cues used by predators (Ghalambor and Martin 2002), and defending their nests when predators attack (Blancher and Robertson 1982; Hogstad 2004). For small songbirds, the importance of nest-site selection is well recognized (reviewed by Lima 2009), which can influence predation patterns observed at natural nests (Schmidt and Whelan 1999a; Latif et al. 2012).
The extent to which small songbirds can influence predation risk following nest initiation is less certain. Parental and nestling activity (e.g., begging) at the nest can attract predators and increase predation risk (Martin et al. 2000), so parents modulate activity at the nest to avoid increasing risk (Ghalambor and Martin 2002; Eggers et al. 2008). Birds can further reduce predation risk by defending their nests, either actively (Blancher and Robertson 1982; Hogstad 2004) or passively (Halupka 1998). Small birds exhibit various defensive behaviors (Ghalambor and Martin 2002; Colombelli-Négrel et al. 2010; see also review by Montgomerie and Weatherhead 1988), but some have doubted the efficacy of such behavior against certain predators (e.g., nocturnal predators; Bradley and Marzluff 2003). Nevertheless, studies do provide evidence for effective nest defense even by small songbirds (initially reviewed by Martin 1992; see also Pietz and Granfors 2008), with intensity and efficacy dependent on food availability (Duncan Rastogi et al. 2006), nest-site quality (Remeš 2005), or predator type (Schmidt and Whelan 2005).
By definition, nest predation involves predators, but determining the extent to which parents are involved can help narrow the range of mechanisms and thus causal factors underlying a pattern of interest. Patterns could arise exclusively from variation in predator ecology, namely their abundance or behavior (Thompson 2007). Parents can adaptively respond to these patterns when selecting nest sites, in which case parents can influence observed patterns (Schmidt and Whelan 1999a; Latif et al. 2012) but leaving predators as the fundamental drivers of predation risk (pathway 1, Fig. 1). Alternatively, post-initiation parental behavior (i.e., nest defense or nest activity) can modulate predation-risk patterns if parental behavior itself varies (pathway 2, Fig. 1), or if parental interactions vary among ecologically different predator species (pathway 3, Fig. 1). If predation patterns are driven exclusively by predator ecology, information regarding alternative prey for predators (Schmidt and Ostfeld 2003a) or predator-habitat relationships (Chalfoun et al. 2002; Schmidt and Ostfeld 2003b) could illuminate underlying mechanisms. Alternatively, if parental behavior modulates observed patterns, food availability for nesting birds (Martin 1992), the presence of conspecifics (Hogstad 1995; Sperry et al. 2008), or factors influencing how parents respond to predators, and vice versa, may also be relevant.
Experimental nests (i.e., artificial nests) provide a potentially useful tool for examining the role of post-initiation parental behavior as a driver of nest predation patterns. Experimental nests have been used widely to study nest predation (reviewed by Major and Kendal 1996), but experimental predation rates and patterns often differ from those experienced by natural nests raising questions about the relevance of experimental-nest data (Faaborg 2006; Moore and Robinson 2004). Among the major reasons suspected for these differences are that experimental nests lack parents (Weidinger 2002). Analysis of the differences in experimental versus natural predation rates and patterns could therefore suggest how parents contribute to predation risk (Weidinger 2002).
We studied the mechanistic pathways underlying predation rates and patterns experienced by a population of Yellow Warblers (Setophaga petechia; Fig. 2) over an 8-year period (2001–2008). Previous work in this study system documented the adaptive significance of nest microhabitat selection for avoiding predation, the principal cause of nest failure. Parents adaptively favored nest-site concealment levels associated with reduced predation risk (Latif et al. 2012), but maladaptively favored microhabitat patch compositions associated with elevated predation risk (Latif et al. 2011). Experimental nests placed in microhabitats also occupied by natural nests recorded similar microhabitat-predation patterns, suggesting predator ecology as the main driver of microhabitat-related predation patterns (pathway 1, Fig. 1). Nest-survival rates were highly variable, suggesting a possible factor contributing to the persistence of maladaptive nest microhabitat preferences; non-microhabitat sources of variability might reduce the contribution of microhabitat-predation patterns (i.e., % variance explained) to overall fecundity and thus reduce the cost of maladaptive nest-site preferences. We therefore expected a closer examination of non-microhabitat correlates of predation rates to provide some context for understanding previous work by further illuminating additional factors contributing to predation risk. Studies elsewhere have identified breeding densities (Schmidt and Whelan 1999b; Hogstad 1995; Perry et al. 2008), seasonal timing, and nest age (Nur et al. 2004, Grant et al. 2005) as potentially important correlates of predation rates, so we were interested in their importance here. Additionally, Yellow Warblers in this system were heavily parasitized by the Brown-headed Cowbird (Molothrus ater; hereafter cowbird), which can affect nest predation in various ways (Arcese et al. 1996; Peer and Bollinger 2000; Tewksbury et al. 2002; Hoover and Robinson 2007), so we were also interested in parasitism relationships with predation risk.
We examined whether parents modulated nest predation patterns related to breeding territory density, seasonal timing, and nest age by comparing patterns observed at natural nests to those observed at experimental nests without parents. We first analyzed patterns across the entire study period to identify those generally experienced by natural nests. We then compared natural patterns to those recorded at experimental nests during 2 years when both were monitored concurrently and across a similar spatial extent. Our analysis accounted for differences between natural and experimental nests other than the presence of parents, allowing us to tease apart potential mechanistic pathways underlying observed patterns (i.e., pathway 1 vs. pathways 2 or 3; Fig. 1). Additionally, we compared overall predation rates to examine the relative influence of parental defense (expected to reduce predation rates for natural nests) versus nest activity (expected to elevate predation rates) in determining natural predation rates. Finally, we analyzed predation relationships with brood parasitism allowing consideration of how cowbirds might affect nest predation risk and patterns.