Oecologia (1999) 120:304310

Springer-Verlag 1999

Jutta C. Burger Michael A. Patten John T. Rotenberry Richard A. Redak

Foraging ecology of the California gnatcatcher deduced from fecal samples

Received: 30 December 1998 / Accepted: 28 April 1999

Abstract The California gnatcatcher is a threatened species essentially restricted to coastal sage scrub habitat in southern California. Its distribution and population dynamics have been studied intensely, but little is known about its diet. We identied arthropod fragments in 33 fecal samples of the California gnatcatcher to gain insight into its foraging ecology and diet. Fecal samples were collected from adult males, adult females, edglings, and nestlings. Leaf- and planthoppers (Homoptera) and spiders (Araneae) predominated numerically in samples. Spider prey was most diverse, with eight families represented. True bugs (Hemiptera) and wasps, bees, and ants (Hymenoptera) were only minor components of the gnatcatcher diet. Gnatcatcher adults selected prey to feed their young that was larger than expected given the distribution of arthropod size available in their environment, and chicks were provisioned with larger prey items and signicantly more grasshoppers and crickets (Orthoptera) and spiders than adults consumed themselves. Both adults and young consumed more sessile than active prey. Further studies are needed to determine whether arthropods sampled in coastal sage scrub that are common in fecal samples are good indicators of California gnatcatcher habitat.

Key words California gnatcatcher Coastal sage scrub Fecal analysis Foraging ecology Polioptila californica

Introduction
What organisms eat is of fundamental importance to their ecology and behavior. Food inuences everything from predator and prey population dynamics and foraging patterns to the niche structure of populations, habitat associations, and competitive relationships. Unfortunately, despite the obvious benets that an understanding of foraging ecology would yield in improving management plans for threatened and endangered birds, we know little about what most birds eat (Rosenberg and Cooper 1990). Thus, our objective was to assess foraging ecology and dietary composition in the U.S. federally threatened California gnatcatcher (Sylviidae: Polioptila c. californica) through identication of prey remains in fecal samples. The California gnatcatcher is endemic to coastal sage scrub shrubland vegetation in cismontane southern California. Its life history and distribution are welldocumented (Woods 1949; Atwood 1988; Rotenberry and Scott 1998), but its dietary preferences are poorly known. Associations have been predicted and observed between arthropod prey abundance and foraging activity (MacArthur and Pianka 1966; Hespenheide 1971), habitat selection (Beaver and Baldwin 1975), population size (Crawford and Jennings 1989), migrant numbers (Griscom 1950), and even vagrant numbers (Patten and Burger 1998) of insectivorous birds. However, whether or not food availability is limiting the California gnatcatcher and aecting its local distribution and/or reproduction is unknown. Arthropod abundance on coastal sage scrub shrub species within gnatcatcher territories may not be correlated with nest success (Roach 1989), but arthropod abundance and vegetation in concert provide a robust predictor for California gnatcatcher presence at a given site (J.T. Rotenberry and R.A. Redak, unpublished data). By

J.C. Burger (8) R.A. Redak Department of Entomology, University of California, Riverside, CA 92521, USA e-mail: jburger@citrus.ucr.edu, Fax: +1-909-7874733 M.A. Patten J.T. Rotenberry Department of Biology, University of California, Riverside, CA 92521, USA J.T. Rotenberry R.A. Redak Center for Conservation Biology, University of California, Riverside, CA 92521, USA

305

dening which arthropods in occupied habitat are actually known prey, we may be able to more easily detect associations between arthropods and birds, and more accurately estimate the importance of arthropod prey in predicting site occupation by California gnatcatchers. We analyzed fecal samples of the California gnatcatcher to (1) estimate the relative contribution of arthropod orders to gnatcatcher diet, (2) deduce size ranges of prey items, (3) compare prey size distributions and prey activity to those of arthropods sampled in coastal sage scrub to determine the degree of selectivity/ opportunism by birds, and (4) compare diets of immature and mature birds. Once we have established which arthropods are consumed by gnatcatchers, we will be better able to identify high versus low food resource sites from arthropod surveys.

Materials and methods

Fecal samples from adult and nestling California gnatcatchers were collected at Marine Corps Air Station Miramar (located about 10 km north of San Diego, Calif., and referred to here as Miramar) during the breeding season of 1995 and 1996 by federally permitted researchers from San Diego State University. Samples were the result of ``in-hand'' defecation during banding operations, either from mist-netted adults and juveniles or from chicks in nests. Fecal samples were collected from 24 June to 25 July in 1995 and from 7 May to 19 July in 1996. In all, we analyzed 33 samples representing birds from 26 nests. Samples were stored in 80% ethanol at the time of collection and later provided to us for analysis. Each sample was centrifuged and treated with about 2 ml of a concentrated solution of NaOH in a hot water bath for 12 h to soften the feces. Samples were subsequently centrifuged, rinsed with distilled water, centrifuged again, and nally transferred and stored in 95% ethanol. Whenever possible, we removed the mucous lining of fecal sacs before softening in order to keep extraneous materials from clouding samples. Arthropod fragments were easily isolated after this treatment. We sampled arthropods using a modied blower-vacuum (Bungton and Redak 1998). Samples were collected from 30 sites in coastal sage scrub vegetation across Miramar from May to early June 1995 and in May 1996, partly overlapping in time with the collection of fecal samples. We considered vacuum sampling to be a reliable means of measuring food availability (see Cooper and Whitmore 1990; Poulin and Lefebvre 1997) because it eectively samples the primarily sessile arthropods that would be encountered by birds while gleaning from shrubs. Reference collections of these arthropods were used to help identify prey items from fragments in samples. In some cases, intact arthropod specimens were macerated in order to match arthropod parts with those found in fecal samples. We identied fragments from samples to the lowest taxonomic level possible. Fragments from larvae of holometabolous insects (primarily moths/butteries and beetles) were lumped together as ``larvae.'' Often, fragments such as mandibles, antennae, legs, wings, and reproductive parts could be characterized for specic orders; in some cases fragments could be recognized as parts from a specic genus or species. When distinct body parts were found that could not be associated with an order, parts were counted as unidentied individuals. The ``miscellaneous'' category presented here refers to minimum estimates of these individuals together with very uncommon groups such as booklice (Psocoptera) and sow bugs (Isopoda). Particles too small and/or nondescript to be associated with an order or to be identied as belonging to an otherwise unrecorded arthropod were disregarded. All fecal samples contained ground (likely animal) matter that was unidentiable and therefore

ignored in analysis. These particles comprised less than half of the volume of each fecal sample. Minimum numbers of arthropods per fecal sample were estimated by summing pairs of matching body parts and counting unique body parts for a given taxon. We estimated prey lengths primarily by averaging the lengths of eld-collected reference specimens in species/taxonomic groups for which fragments were found. Because specic reference specimens of adult Hymenoptera and Lepidoptera (moths) and larvae could not be identied in feces, their lengths were assigned as the average length of all eld-collected specimens of their group. Grasshoppers (Acrididae), were represented in feces by a range of sizes (from nymphs to adults) that could be directly estimated using allometric equations. We measured the lengths of two heavily sclerotized body parts commonly found in samples (crescent of hind femur and mandible) for 28 reference Orthoptera of known body length. Fragment lengths (in mm) were related to total body lengths by exponential functions [crescent (c): total length = 3.54e0.68c, R2 = 0.95; mandible (m): total length = 3.33e1.08m, R2 = 0.95]. The biomass of all orders was calculated using previously published functions relating body length to dry mass (Rogers et al. 1977). The relative contribution of various arthropod orders to gnatcatcher diet was expressed both in terms of percent by number of individuals and by estimated biomass. Frequency distributions of prey lengths were compared with the distribution of arthropod lengths from vacuum samples at Miramar using a Kolmogorov-Smirnov test for goodness of t (Sokal and Rohlf 1981, p. 716). Only the most common arthropod orders found in feces (Araneae, Coleoptera, Homoptera, Orthoptera) were included in comparisons of lengths. Relative numbers of active (Diptera, Hymenoptera, Lepidoptera) and sessile (all others) arthropods found in gnatcatcher feces were compared with those collected in the eld using a chi-square test for goodness of t. We compared diets among age classes using multivariate analysis of variance (MANOVA). Individual means were compared with a Tukey honest signicant distance (HSD) test (SAS 1996).

Results
Arthropods of ten orders were found in fecal samples. Diagnostic fragments of each could be used for identication (Table 1, Fig. 1). Homoptera were not only the most consistently found component of feces (Fig. 2), but also the most abundant (Table 2). However, most Homoptera were small and therefore did not make up the bulk of biomass consumed (Table 2). Spiders (Araneae) were found in most samples (Fig. 2). The minimum number of spiders in feces could easily be quantied by counting pairs of matching chelicerae, fangs, or individual epigyna (reproductive structure of female). They, along with Coleoptera, made up a large component of diet both in terms of number of individuals and in terms of biomass consumed (Table 2). Adult Lepidoptera, although relatively common in samples, could not be quantied easily, as they were usually identied only by the presence of scales. Assuming that in most instances they occurred as a single individual in a sample, Lepidoptera represented only a small fraction (<6%) of the diet, both in terms of number of individuals and biomass. Orthoptera were found in roughly half of the samples analyzed (Fig. 2); however, when their contribution to diet was analyzed in terms of biomass, it outweighed that of all other orders (Table 2). Nearly all samples also contained a few fragments (``miscellaneous'') that could not be further matched or identied.

306 Table 1 Types of arthropod fragments found in California gnatcatcher feces. See Fig. 1 for appearance of diagnostic fragments Order/group Araneae Coleoptera Diptera Hemiptera Homoptera Hymenoptera Isopoda Larvae Lepidoptera Orthoptera Psocoptera Fragments found Chelicerae, fangs, epigyna, palpal bulbs, fragments, leg fragments, eyes, spinnerettes Elytra fragments, leg segments, tarsomeres, antennal segments, exoskeletal fragments, heads, head fragments, mandibles Wings or wing fragments, head fragments, antennal segments, tarsomeres, tarsal claws Leg segments, heads Heads, head fragments, exoskeletal fragments, wings, leg fragments, tarsomeres Head, wing fragments, propodeum, mandible Antennal and leg segments, body fragments Mandibles, head capsules, skins, prolegs Scales, heads, legs, antennae Mandibles, lacinia, pronotum, wing fragments, leg fragments, crescent of hind femur Head

Several arthropod prey items could be identied to family or further (Table 3). Araneae were the most diverse, representing 11 known species and 8 families (most commonly Salticidae). Coleoptera families primarily represented were Chrysomelidae and Curculionidae. Homoptera were represented by a minimum of 8 species from 5 families [predominantly Tiaja sp. (Cicadellidae), Oenopterix sp., and Dictyssa spp. (Issidae)]. Orthoptera consisted mostly of grasshoppers (Acrididae), although a few tree crickets (Gryllidae) were found as well. Prey length estimates ranged from 1.1 to 14.5 mm, with average prey length at 4.3 mm (Table 4, Fig. 3). Prey biomass ranged from 0.1 to 40.1 mg dry weight, with most commonly consumed arthropods (Dictyssa spp. and Tiaja sp.) weighing between 28 mg. Orthoptera were by far the largest prey items consumed. Size distributions of prey in chick and edgling feces were signicantly different from size distributions of the same orders collected in the eld (Fig. 3; Kolmogorov-Smirnov goodness of t for chicks D52 = 0.5519, P < 0.01, for edglings D52 = 0.6081, P < 0.01). Both chicks and edglings consumed prey that were larger than the most abundant items in their surroundings. Prey size distribution in adult male and female gnatcatchers appeared not to vary signicantly from that of eld-collected arthropods. Gnatcatchers included proportionally more sessile arthropods (85.2%) in their diet than were available (70.9%) in their environment (v12 = 29.85, P < 0.001). Age was a signicant factor inuencing the diet of the California gnatcatcher (Fig. 4; Wilks' L12,69 = 0.29, P < 0.001). Chicks were provisioned with signicantly more Araneae and Orthoptera than adults consumed. Adults fed upon signicantly more Homoptera than did immatures. Coleoptera were consumed equally across all ages. Males and females had similar diets. Juveniles appeared to have a diet intermediate to that of adults and chicks, consuming fewer Homoptera than did adults but also eating fewer Orthoptera than did chicks.

blue-gray (P. caerulea) gnatcatchers, are gleaners that forage by picking arthropods o foliage (Atwood 1988); they occasionally obtain winged prey in ight (Root 1967). Only two studies have investigated the diet of gnatcatchers. A cursory analysis of California and bluegray gnatcatcher stomachs of ambiguous age, date, and location of collection conducted early in this century reported that the two species had similar diets and fed primarily on Homoptera, followed in order of importance by adult Hymenoptera and Coleoptera (Beal 1907). Larvae (primarily of Lepidoptera and Coleoptera) comprised only 5% of the diet and Araneae (spiders) even less. A more recent analysis of blue-gray gnatcatcher diet and feeding ecology in deciduous oak scrub-woodland in northern California included thorough stomach analyses of both young and adult birds (Root 1967). Homoptera (specically Membracid planthoppers) again comprised the majority of gnatcatcher food, although the relative contribution of other orders diered from that found by Beal (1907). Dietary composition of California gnatcatchers was similar to that of the blue-gray gnatcatchers studied by Root (1967). Homoptera were a major component of adult California gnatcatcher diet, chicks consumed signicantly more Araneae and Orthoptera than did adults, and average prey size for chicks was signicantly larger than that for adults. Membracid planthoppers (Homoptera) were a major component of blue-gray gnatcatcher diet (Root 1967), but were absent as prey in our study despite their availability. Instead, other small,
c Fig. 1 Diagnostic arthropod fragments found in California gnatcatcher feces (described clockwise beginning at the upper left-hand corner of each gure panel). A Homoptera (Issidae: Dictyssa obliqua): entire wing, hind tibia fragment, wing fragment, hind tibia fragment. B Homoptera (Cicadellidae: Taija sp.): wing fragment, hind tibia fragment, head fragment with eye, clypeus/clypellus of face. C Hemiptera: head, femur. D Orthoptera: mandible (Gryllidae), tibial spine, tibial joint, mandible, pronotal plate, and hind femoral crescent (Acrididae). E Araneae: chelicera with fang, eye (Salticidae), leg fragment, female epigynum, fang. F Diptera: wing (suborder Nematocera), fragmented tarsus (tarsomeres and claw likely from Asilidae), intact tarsus. G Coleoptera: elytra fragment (Mordellidae), elytra fragment and entire leg (Chrysomelidae). H Larvae of Coleoptera and Lepidoptera: mandible, proleg, large mandible, small mandible

Discussion
The California gnatcatcher and its nearest relatives in the United States, the black-tailed (P. melanura) and

307

308

heavily sclerotized Homoptera (Issidae and Cicadellidae) were abundant. Indeed, an earlier diet analysis of birds in the context of the fruit industry in California identied more Homoptera in the gnatcatcher diet than for any other bird surveyed (Beal 1907). In contrast to previous studies of gnatcatcher diet (Beal 1907; Root 1967), Hymenoptera were a small component of the diet of California gnatcatchers. Adult California gnatcatchers chose larger prey items for chicks than for their own consumption. Previous studies have also found evidence that average prey size is larger for young birds (Root 1967; Bull et al. 1992; Kaspari and Joern 1993) and immatures of other taxa (Sempeski and Gaudin 1996) than it is for adults. It is possible that this pattern is an artifact of dierences in digestion abilities of young and adult birds. Although there have been no specic studies on the ontogeny of digestion in gnatcatchers, digestive eciency is independent of age in the red junglefowl (Gallus gallus; Jackson and Diamond 1995) and is likely to be so in all birds given their rapid maturation. We believe this phenomenon is best explained in terms of costs and benets to the adults. When risk of predation is high and is inuenced by activity around a nest (e.g., coming and going of adults, begging from chicks), birds that bring a few large prey items to their nest should have greater tness than birds that bring more small prey items. The vast majority of nest failures experienced by California gnatcatchers are the result of predation (Braden et al. 1997; Sockman 1997). Furthermore, activity around the nest, even if by human investigators, increases the risk of predation (Skutch 1949, 1985; Haskell 1994; Leech and Leonard 1997; Sockman 1997; cf. Roper and Goldstein 1997). Our results are also consistent with the central-place foraging theory (Orians and Pearson 1979), which predicts that organisms which bring prey to a central site for consumption (as adults do for chicks) will choose larger

prey due to the energetic costs associated with travel to and from that location. Although adult gnatcatchers disproportionately selected large prey when feeding chicks, their own diet
Table 2 Overall diet composition of California gnatcatchers based on the number of arthropods found in fecal samples and their biomass (estimated using allometric equations; see text) Order Proportion of diet by number of individuals (%) 25 18 13 12 8 7 6 6 3 2 Estimated proportion of diet by biomass (%) 9 18 8 15 37 1 7 3 2 <1

Homoptera Araneae Miscellaneous Coleoptera Orthoptera Larvae Diptera Lepidoptera Hemiptera Hymenoptera

Table 3 Species identied from most common orders in fecal samples. Names in parentheses are tentative Order Araneae Families (Anyphaenidae) Clubionidae Gnaphosidae Oxyopidae Philodromidae Salticidae Theridiidae Thomisidae Anthribidae Buprestidae Chrysomelidae Curculionidae Mordellidae Diptera Hemiptera Homoptera Scarabaeidae Asilidae Chironomidae (Sciaridae) Lygaeidae Nabidae Aphididae Cicadellidae Cixiidae Dictyopharidae Issidae Hymenoptera Braconidae/ Ichneumonidae Formicidae Acrididae Gryllidae Species Unknown Cheiracanthium inclusum Sergiolus montanus Hamataliwa grisea Oxyopes salticus Peucetia sp. Tibellus chamberlini Habronattus hirsutus Salticus palpalis Sassacus (papenhoei) Theridion intervallatum Xysticus sp. Unknown Unknown Chlamisus sp. Other unknowns Sittonia californicus Mordellis sp. Other unknowns Dichelonyx sp. Unknown Unknown Unknown Nyssus sp. Unknown Unknown Homalodisca lacerta Huleria quadripunctata Tiaja sp. Unknown Unknown Dictyssa obliqua Oenopterix manea Unknown Unknown Unknown Oecanthus sp.

Coleoptera

Fig. 2 Frequency of detection of arthropod orders, miscellaneous unidentied arthropods, and larvae in California gnatcatcher feces (n = 33 samples)

Orthoptera

309 Table 4 Estimated size and biomass ranges of California gnatcatcher prey. Size ranges (dry lengths) were obtained from reference specimens of the same species and/or similar species as found in gnatcatcher scat. Biomass was calculated using formulas referred to in the text. Values with asterisks represent averages obtained from broader taxonomic groups in reference collections that were found in scat. Values in parentheses represent averages from other arthropods identied from scat when no better estimate could be obtained Order Araneae Coleoptera Diptera Hemiptera Homoptera Hymenoptera Lepidoptera Miscellaneous Orthoptera Larvae Size range (mm) 1.18.0 2.911.19 1.9*7.6* 2.15.3* 1.27.4 2.0*3.6* (4.3) (4.3) 4.814.5 4.5* Biomass range (mg) 0.119.8 0.6026.52 0.18*6.4* 0.32.2* 0.18.1 0.1*0.4* (1.2) (1.3) 2.540.1 0.4*

Fig. 4 Average (+SE) number of prey of the most common arthropod orders in the diet of chick (n = 11), edgling (n = 8), adult male (n = 7), and adult female (n = 7) California gnatcatchers. Dierent letters represent signicant dierences across ages for a given prey type (Tukey's HSD, a < 0.05). No signicant dierences were found across ages for Coleoptera

Fig. 3 Size distribution of Araneae, Coleoptera, Homoptera, and Orthoptera prey items found in gnatcatcher scat compared to size distribution of these arthropod orders collected in the eld (n = 5422) for chicks (52 prey items, n = 11 birds), edglings (25, n = 8), females (46, n = 7), and males (31, n = 7). Arthropod lengths from chicks and edglings were signicantly dierent from eld-collected arthropods (P < 0.01, Kolmogorov-Smirnov goodness of t), but males and females were not (P > 0.10)

reected a more opportunistic foraging behavior in which prey did not dier in size from the arthropods available in the eld (Fig. 3). Observed dierences in both taxonomic composition and size of prey for chicks are likely a reection of changes in foraging tactics by adults. Similarly, adult blue-gray gnatcatchers forage more actively, engage in more aerial attacks, and utilize the herb layer more when feeding their young (Root 1967). Gnatcatchers are best described as ``near-surface searchers'' that glean arthropods o foliage while moving quickly through substrate (Root 1967; Robinson and Holmes 1982). We found that California gnatcatchers

took proportionately more sessile prey (such as leafhoppers, beetles, and spiders) than were available in the environment. These results suggest that gnatcatchers take advantage of high prey vulnerability and are not purely opportunistic in terms of prey activity level, another nding consistent with previous studies in various taxa (Onkonburi and Formanowicz 1997; Tikkanen et al. 1997). Therefore, availability of both large prey and highly vulnerable prey should be considered when evaluating the quality of gnatcatcher habitat. For example, although Hymenoptera (which are primarily small and active) were the second most abundant order collected in the eld, they were essentially ignored by foraging gnatcatchers. There are drawbacks to assessing diet and foraging ecology based on an analysis of fecal samples (Hartley 1948; Ralph et al. 1985). Drawbacks include the underestimation of soft-bodied arthropods and arthropods that lack distinctive sclerotized parts. Nonetheless, numerous studies have successfully used fecal analysis (e.g., Calver and Wooller 1981; Moeed and Fitzgerald 1982; Ralph et al. 1985; Moreby 1988). Samples can be collected easily from captured birds, or, in the case of larger birds, from areas subtending perches or nests (Elliott and Cowan 1983; Smith and Calver 1984; Bull et al. 1992). Arthropod fragments are smaller and more dicult to identify than using other methods of diet analysis, but sample collection results in virtually no additional stress to the bird and ensures that the arthropods identied were not only ingested but also digested. We chose fecal analysis to study California gnatcatcher diet because it remains virtually the only acceptable method for studying diets of endangered or threatened species. We likely underestimated the number of individuals in a particular fecal sample as well as the

310

number of soft-bodied insects (e.g., adult Lepidoptera) and larvae of holometabolous insect orders; however, our analysis does provide a necessary rst step in characterizing the diet and foraging ecology of this threatened species. Further studies associating arthropods with the occurrence of this species may be improved by including only the known food sources and/or known size ranges in analyses. These studies may enable us to better understand why California gnatcatchers are not found in some coastal sage scrub areas that appear, supercially, to be suitable habitat. Furthermore, we can learn more about prey selectivity by the California gnatcatcher by comparing fecal samples from other insectivorous birds foraging in the same habitats.
Acknowledgements Ahrash Bissell and Kylie Fischer from Don Hunsaker's research group at San Diego State University graciously took time to collect fecal samples from Miramar. Roger Burks gave insight and advice into identication of arthropod fragments. Thomas R. Prentice impressed us by identifying minute Araneae fragments to family and even species. Kathleen A. Campbell veried identication of Hymenoptera. This research was partially funded by the Department of Defense Legacy Resources Management Program under Cooperative Agreement no. N68711-94-LT-4042 to R.A.R. for arthropod surveys at Marine Corps Air Station Miramar and Marine Corps Base Camp Pendleton.

References
Atwood JL (1988) Speciation and geographic variation in blacktailed gnatcatchers. Ornithol Monogr 42 Beal FEL (1907) Birds of California in relation to the fruit industry, part 1. US Dept Agric Biol Surv Bull 30:8486 Beaver DL, Baldwin PH (1975) Ecological overlap and the problem of competition and sympatry in the western and Hammond's ycatchers. Condor 77:113 Braden GT, McKernan RL, Powell SM (1997) Eects of nest parasitism by the brown-headed cowbird on nesting success of the California gnatcatcher. Condor 99:858865 Bungton ML, Redak RA (1998) A comparison of vacuum sampling versus sweep-netting for arthropod biodiversity measurements in California coastal sage scrub. J Insect Conserv 2:99106 Bull EL, Beckwith RC, Holthausen RS (1992) Arthropod diet of pileated woodpeckers in northeastern Oregon. Northwest Nat 73:4245 Calver MC, Wooller RD (1981) Seasonal dierences in the diets of small birds in the Karri Forest understory. Aust Wildl Res 8:653657 Cooper RJ, Whitmore RC (1990) Arthropod sampling methods in ornithology. Stud Avian Biol 13:2937 Crawford HS, Jennings DT (1989) Predation by birds on spruce budworm Chorisoneura fumiferana: functional, numerical, and total responses. Ecology 70:152163 Elliott CL, Cowan CA (1983) The food habits of an American kestrel (Falco sparverius) in interior Alaska. Murrelet 64:6364 Griscom L (1950) Northeastern maritime region, fall migration. Audubon Field Notes 4:45 Hartley PHT (1948) The assessment of the food of birds. Ibis 90:361379 Haskell D (1994) Experimental evidence that nestling begging behaviour incurs a cost due to nest predation. Proc R Soc Lond B 257:161164 Hespenheide HA (1971) Food preference and the extent of overlap in some insectivorous birds, with special reference to the Tyrannidae. Ibis 113:5972

Jackson S, Diamond J (1995) Ontogenetic development of gut function, growth, and metabolism in a wild bird, the red jungle fowl. Am J Physiol 269:11631173 Kaspari M, Joern A (1993) Prey choice by three insectivorous grassland birds: reevaluating opportunism. Oikos 68:414430 Leech SM, Leonard ML (1997) Begging and the risk of predation in nestling birds. Behav Ecol 8:644646 MacArthur RH, Pianka ER (1966) On optimal use of a patchy environment. Am Nat 100:603609 Moeed A, Fitzgerald BM (1982) Foods of insectivorous birds in forest of the Orongorongo Valley, Wellington, New Zealand. N Z J Zool 9:391402 Moreby SJ (1988) An aid to the identication of arthropod fragments in the feces of gamebird chicks (Galliformes). Ibis 130:519526 Onkonburi J, Formanowicz DR Jr (1997) Prey choice by predators: eect of prey vulnerability. Ethol Ecol Evol 9:1925 Orians GH, Pearson NE (1979) On the theory of central place foraging. In: Horn DJ, Mitchell RD, Stairs GR (eds) Analysis of ecological systems. Ohio State University Press, Columbus, pp 154177 Patten MA, Burger JC (1998) Spruce budworm outbreaks and the incidence of vagrancy in eastern North American wood-warblers. Can J Zool 76:433439 Poulin B, Lefebvre G (1997) Estimation of arthropods available to birds: eect of trapping technique, prey distribution, and bird diet. J Field Ornithol 68:426442 Ralph CP, Nagata SE, Ralph CJ (1985) Analysis of droppings to describe diets of small birds. J Field Ornithol 56:165174 Roach JD (1989) The inuence of vegetation structure and arthropod abundance on the reproductive success of California black-tailed gnatcatchers, Polioptila melanura californica. MSc thesis, San Diego State University Robinson SK, Holmes RT (1982) Foraging behavior of forest birds: the relationships among search tactics, diet, and habitat structure. Ecology 63:19181931 Rogers LE, Buschbom RL, Watson CR (1977) Length-weight relationships of shrub-steppe invertebrates. Ann Entomol Soc Am 70:5153 Root RB (1967) The niche exploitation pattern of the blue-gray gnatcatcher. Ecol Monogr 37:315350 Roper JJ, Goldstein RR (1997) A test of the Skutch hypothesis: does activity at nests increase nest predation risk. J Avian Biol 28:111116 Rosenberg KV, Cooper RJ (1990) Approaches to avian diet analysis. Stud Avian Biol 13:8089 Rotenberry JT, Scott TA (1998) Biology of the California gnatcatcher: lling in the gaps. West Birds 29:237241 SAS (1996) SAS statistical software, version 6.12. SAS Institute, Cary, NC Sempeski P, Gaudin P (1996) Size-related shift in feeding strategy and prey-size selection in young grayling (Thymallus thymallus). Can J Zool 74:15971603 Skutch AF (1949) Do tropical birds rear as many young as they can nourish? Ibis 91:433450 Skutch AF (1985) Clutch size, nest success, and predation on nests of Neotropical birds: a review. Ornithol Monogr 36:575 594 Smith GT, Calver MC (1984) The diet of the nestling noisy scrubbird, Atrichornis clamosus. Aust Wildl Res 11:553558 Sockman KW (1997) Variation of life-history traits and nest-site selection aects risk of nest predation in the California gnatcatcher. Auk 114:324332 Sokal RR, Rohlf FJ (1981) Biometry, 2nd edn. Freeman, New York Tikkanen P, Muotka T, Huhta A, Juntunen A (1997) The roles of active predator choice and prey vulnerability in determining the diet of predatory stoney (Plecoptera) nymphs. J Anim Ecol 66:3648 Woods RS (1949) Polioptila melanura californica Brewster: blacktailed gnatcatcher. In: Bent AC (ed) Life histories of North American thrushes, kinglets, and their allies. US Natl Mus Bull 196:374381