Habitat and Nesting Success of Blue Jays (cyanocitta Cristata): Importance of Scale

—Nesting success of birds often is influenced by habitat features surrounding nests. However, few studies have investigated habitat influences at multiple levels and spatial scales. We examined the relationship between Blue Jay (Cyanocitta cristata) nesting success and habitat at three levels (macrohabitat, microhabitat, and nest site) and at three different scales within the microhabitat level (3.1, 0.3, and 0.07 ha circular plots surrounding nests) in central Florida. We defined successful nests as those fledging at least one offspring. An eight-fold difference in probability of nesting success existed among seven distinct macrohabitats, with Mayfield success estimates ranging from 8 to 63%. However, no difference in number of fledglings per nest existed among macrohabitats when unsuccessful nests were excluded from the analysis. Macrohabitats in which nests were more often successful had more extensive slash pine (Pinus elliottii) forest. Logistic regression analyses indicated that slash pine canopy cover was significantly related to nesting success at all three microhabitat scales, although the tree species in which a nest was placed was unimportant to nesting success. Other microhabitat and nest-site features were unrelated to nesting success. Strong circumstantial evidence indicates predation was the major cause of nest failure, but relative abundance of one potential predator, the gray squirrel (Sciurus carolinensis), was not related to nesting success in macrohabitats or slash pine canopy cover within microhabitats. In this study, logistic regression models using habitat features measured at higher levels and larger scales had better predictive power, indicating that habitat may have influenced predation at scales larger than that of a typical nest patch. Received 21 August 2001, accepted 8 June 2002. RESUMEN.—El éxito de nidificación en las aves es a menudo influenciado por las caracterı́sticas del hábitat que circunda a los nidos. Sin embargo, pocos estudios han investigado las influencias del hábitat a múltiples niveles y escalas espaciales. Examinamos la relación entre el éxito reproductivo de Cyanocitta cristata y el hábitat a tres niveles (macrohábitat, microhábitat y sitio del nido) y a tres niveles diferentes dentro del microhábitat (parcelas circulares alrededor de los nidos de 3.1, 0.3 y 0.07 ha) en Florida central. Definimos como nidos exitosos a aquellos en que por lo menos una cria llegó a la etapa de emplumamiento. Encontramos una diferencia de ocho veces en la probabilidad de anidar exitosamente entre siete macrohábitats distintos, con valores de la estimación del éxito de Mayfield entre un 8 a un 63%. Sin embargo, no se encontró diferencia en el número de volantones por nido entre macrohábitats cuando se excluyeron los nidos no exitosos del análisis. Los macrohábitats en que los nidos fueron exitosos más a menudo, presentaban una mayor extensión de bosques de pino (Pinus elliottii). Los análisis de regresión logı́stica indicaron que la cubierta del dosel de P. elliottii se relacionó significativamente con el éxito de nidificación a los tres niveles de microhábitat, aunque la especie del árbol en que se encontraba un nido no fue importante para el éxito de nidificación. Otras caracterı́sticas del microhábitat y rasgos del sitio del nido no se relacionaron con éxito de nidificación. Fuerte evidencia circunstancial indica que la depredación fue la mayor causa de fracaso de los nidos, pero la abundancia relativa de un de predador potencial, la ardilla gris (Sciurus carolinensis), no se relacionó con el éxito de nidificación en los macrohábitats ni con la cubierta del dosel de P. elliottii dentro de los microhábitats. En este estudio, los modelos de regresión logı́stica que utilizan caracterı́sticas del hábitat medidas a niveles más altos y a mayores escalas tuvieron un mejor poder de 4 Present address: Department of Biology, Oberlin College, Oberlin, Ohio 44074, USA. E-mail: keith. [email protected] 972 [Auk, Vol. 119 TARVIN AND GARVIN predicción, indicando que el hábitat puede influenciar a los patrones de depredación a una escala mayor que a la de un parche tı́pico de un nido. SUCCESS OF OPEN passerine nests may be influenced by habitat surrounding them (e.g. Martin and Roper 1988, Kelly 1993, Tarvin and Smith 1995). For example, at a coarse-grained level based on community composition (here defined as ‘‘macrohabitat’’), the abundance or distribution of food or predator species may be greatly influenced by characteristics such as abiotic factors, relative abundance of prey species, community structure, etc. At a finergrained, local level (‘‘microhabitat’’), structural complexity (e.g. plant architecture, stem density) of surrounding features may influence the ability of predators to find nests by interfering with visual or olfactory cues (Martin and Roper 1988), impeding predator movement through an area (Bowman and Harris 1980), or influencing reward rates when predators must search many similar potential nest sites to find dispersed nests (Martin 1988, Martin and Roper 1988, Tarvin and Smith 1995). A third level at which habitat features may influence nesting success is that of the nest site, which comprises the substrate in or on which the nest is placed and its qualities (e.g. height, concealment, stability, etc.). Features of nest sites may preclude predators from accessing nests (Cullen 1957, Lawton and Lawton 1980), affect the deterrent or vigilant behaviors of parent birds (Belles-Isles and Picman 1986), or simply hide nests from predators (Kelly 1993). Nesting habitat also can influence the probability of nest failure through factors other than nest predation. For instance, structural characteristics of nest sites may affect the probability of nest safety during strong winds, or provide shelter from climatic factors such as rain, snow, insolation, or thermal variation (Walsberg 1981, Marzluff 1988). Within the level of microhabitat, the scale at which habitat influences nesting success may vary, depending on the local suite of predators, distribution of food, or other factors (Martin and Roper 1988, Martin 1992). For example, influence of structural complexity on predator olfactory search efficiency may be important within a certain distance from the nest, but may have little effect at greater distances. Determining the appropriate spatial scale at which to study the relationship between habitat and nesting success often is difficult, and any chosen scale is to some degree arbitrary. Because processes influencing nesting success are likely to be scale dependent, the ability of studies to detect relationships between habitat and nesting success is in part dependent on the scale chosen for measurements. In this study, fate of Blue Jay (Cyanocitta cristata) nests in central Florida was determined and tested for relationships between nesting success and habitat features at the levels of macrohabitat, microhabitat, and nest site. For a subset of nests, we further tested for relationships between nesting success and microhabitat features measured at three different spatial scales. Some published studies have used the term ‘‘nest patch’’ instead of microhabitat (Martin and Roper 1988, Martin 1992, Tarvin and Smith 1995); however, we use the latter term because our study focuses on habitat occurring within arbitrary plot boundaries rather than distinct clumps of vegetation (see Burhans and Thompson 1999). Blue Jays serve as a good species for studies of this kind because they exhibit extreme variation in use of nesting habitat. Blue Jays nest in deciduous, mixed coniferous–deciduous, and coniferous forest, as well as suburban, agricultural, and other human-disturbed landscapes (Tarvin and Woolfenden 1999). Nests range from ;2 m above ground in shrubs or saplings to the uppermost branches of the tallest trees available, and may be placed in virtually any tree species present (Tarvin and Woolfenden 1999). Furthermore, Blue Jays nest in all parts of trees, including crotches in the main trunk, large horizontal limbs, and near the tips of terminal branches. Thus, they nest in a wide array of macrohabitats, as well as a wide range of microhabitats and nest sites within those macrohabitats. If the influence of habitat on Blue Jay nesting success is strong, such relationships should be detectable if an appropriate scale of habitat measurement is used.