Use of coastal beaches by nonbreeding shorebirds along the Pacific coast of North America is poorly understood because survey efforts have targeted large bays and wetlands where birds concentrate. Consequently, we studied coastal shorebird distributions and community composition by conducting monthly surveys at 40, 0.5-km linear plots along sandy beaches of coastal northern California from January through April 1996. We recorded 12 species, whose ranked abundances were consistent between winter (January–February) and spring (March–April). However, five of the six commonest species were more abundant in winter than spring. Among beaches, we detected a mean of 2.6 ± 2.3 (SD; range: 0–9) species and 12.0 ± 14.1 (range: 0–62) individuals. Number of species correlated positively with total shorebird abundance. Most species' abundances varied greatly among sites; nine taxa had aggregated spatial distributions, one species was hyper-dispersed, and two species were randomly distributed among beaches. Species composition of the shorebird community was highly nested, partly because abundant species were more widespread in their occurrence than less-numerous taxa. More species and individuals occurred at sites near Humboldt Bay, probably because high tides forced birds off their principal feeding areas within the bay to adjacent beaches. Variation in shorebird use of beaches suggests that some areas are more important to wintering and migrating shorebirds than other locations. Although current levels of human activity in coastal northern California do not appear to influence distributions of nonbreeding shorebirds, we suggest that monitoring and management plans be considered to minimize potential negative impacts of recreational activities, which are certain to increase with human population size and recreational use of beaches.

Egg type is one factor that is likely to influence the results from artificial nest experiments. Here I document predation rates on Japanese Quail (Coturnix japonica) and the much smaller Zebra Finch (Taeniopygia guttata) eggs on the edge and interior of a tropical forest in Costa Rica. Predation rates on finch eggs were the same on the edge and in the interior of the forest (60.9 and 60.0%, respectively), while predation rates on quail eggs were significantly different on the edge and in the interior (76.5 and 35.0%, respectively). Hence, results from artificial nest experiments may be highly dependent on the type of egg used and caution is advised in their interpretation.

Farallon de Medinilla is a small island in the western Pacific that is leased by the U.S. military from the Commonwealth of the Northern Mariana Islands as an impact area for training exercises. We spent 5.5 h on the island surveying its avifauna on 4 Nov. 1996 and observed a total of 17 species including seabirds, migratory shorebirds, and resident landbirds. Of special interest was the first island record of the endangered Micronesian Megapode (Megapodius laperouse) and breeding colonies of Great Frigatebirds (Fregata minor) and Masked Boobies (Sula dactylatra). Locations of individual megapodes and seabird colonies were mapped and breeding information recorded. Historical records and photographs indicate that since the island has been used as an impact area its vegetative structure has changed from a medium-height, relatively closed canopy forest, to primarily open areas with intermittent patches of low forest. Alterations of native vegetation likely resulted in changes in the density, distribution, and species composition of the island's avifauna, although these changes are difficult to quantify.

Common Tern (Sterna hirundo) nests were both clumped (high density) and solitary (low density) at Windermere Basin, Lake Ontario in 1992. Our objective was to measure egg fate and chick survival up to 14–16 d of pairs nesting in these two dispersion patterns. We first measured the distance to nearest neighbor for 100 randomly selected clutches, and defined low-density nests as those at a distance greater than 2 SD (m) away from the nearest clutch. Two sites contained nests at high density (HD1 & HD2), while low density nests (LD) were dispersed throughout the study site. We randomly selected 20, three-egg clutches in each of the three groups. Nearest neighbor distances of HD1 (1.08 ± 0.71 m) and HD2 (2.09 ± 1.03 m) clutches were similar and significantly different from nearest neighbor distances of LD clutches (21.9 ± 37.3 m). There were no differences in clutch initiation dates or C-egg volumes among the 60 clutches of the three study groups. LD and HD1 clutches experienced heavy egg and subsequent chick loss, while losses of eggs and chicks at HD2 were negligible. The evidence for type of predator was largely circumstantial, but for eggs that were destroyed and chicks that disappeared, the loss patterns suggested primary predation by Ruddy Turnstones (Arenaria interpres) and Black-crowned Night-Herons (Nycticorax nycticorax), respectively. The locations and timing of egg and chick losses indicates that opportunistic behavior of these avian predators nullifies any predatory avoidance advantages of high-density nesting once the predator(s) identify the food resource.

This study investigated the accuracy of 50-m fixed-radius, 100-m fixed-radius, and variable circular-plot point counts to estimate the actual density of breeding Cerulean Warblers (Dendroica cerulea) during the 1997 and 1998 breeding seasons, in Ontario, Canada. Density estimates were compared to actual densities as measured from intensive field observation of pairing and nesting behavior. Estimates of density from each of the techniques were positively correlated with actual density in both years. Both the technique used to census a population as well as the actual density of the population itself affected the accuracy of the derived density estimations. In both years, the 50-m fixed-radius technique overestimated density. In contrast, the 100-m fixed-radius technique and variable circular-plot technique underestimated density; the degree of the underestimate of the 100-m fixed-radius increased as actual density increased. There was no correlation between the degree of underestimation and actual density for the variable circular-plot technique. Although all three methodologies provide relative measures of density, the variable circular-plot technique provides the best absolute assessment of Cerulean Warbler density and is considered most suitable for broad-scale surveys.

We describe a technique for setting mist nets horizontally from canopy platforms, detailing materials needed, installation, and use of the net system. Using a control cord, the operator pulls the net outward from the platform along a cable, and rolls it inward around a manually-operated spindle to remove captured birds. Because nets are not raised from the ground, disturbance to vegetation is minimized. Set-up time averaged 5 min/net after climbing to the platform, and we used a maximum of two nets from a platform. High winds (>40 km/h) and light rain closed canopy nets more often than ground nets, and canopy nets were less effective before leaf-out in spring. In 1751 net-hours, 121 birds of 25 species were captured, and 44 birds (27%) were recorded escaping the nets. The method seeks to parallel conventional netting so as to allow comparisons between strata, making this technique useful for studies that require simultaneous mist-netting in forest canopy and understory levels.

The bird assemblage of a southern temperate rainforest on Chiloé Island, southern Chile, was studied during the breeding season of 1993–1994. The modified variable-circular plot method was used to examine the effects on the estimated overall density and species richness of birds when the size of the plot, duration at the plot, and number of plots were varied, one variable at a time. A total of 459 individuals of 24 species was recorded on 18 plots sampled for a maximum of 10 min with unlimited distance. When increasing either the plot size, sampling length, or number of plots, the number of species detected increased at a decreasing rate, but without reaching an asymptote. The estimated density decreased with increasing plot size, increased with sampling length, and did not change with increasing plot number. However, the precision substantially improved. Plots of 40-m radius counted for 10 min detected 78% of the species found in larger plots and rendered an overall bird density of 34 individuals/ha, almost twice as many as the only comparable study found. On average, more than 75% of all species detected were within 40 m of the observers, using a 4-min sampling period at five plots. The migrant omnivorous flycatcher Elaenia albiceps was the most abundant species (22% of total counts), followed by the hummingbird Sephanoides galeritus, the tapaculo Scelorchilus rubecola, the treecreeper Aphrastura spinicauda, and the parakeet Enicognathus ferrugineus. These species, which were the most common in other studies conducted in similar temperate South American forests, showed species-specific responses to changes in the sampling variables.

I provide a detailed description of an arthropod sampling technique, “branch clipping,” and evaluate its efficacy in measuring food availability for three foliage-gleaning migratory warbler species wintering in Jamaica. I incorporated foraging observations into the sampling protocol to facilitate a match between the distribution of branch-clip samples and the distribution of warbler foraging attacks over available microhabitats. Where the match was imprecise, I weighted branch clip samples accordingly. Foraging observations indicated that most foraging maneuvers used by the warblers were directed at prey located on foliage, which are likely to be well sampled by branch clipping. Published diet analyses permitted the elimination of prey taxa and sizes that were not eaten by the warblers, and suggested that nearly all of the items eaten were captured by branch clipping. Habitat-specific warbler attack rates, which can be used as independent evaluations of food availability measurements, closely matched prey density as estimated by branch clips across three habitat types (citrus orchard, coffee plantation, and dry limestone forest). These data suggest that branch clipping effectively measured food availability for foliage-gleaning warblers in Jamaica. The method is probably also well suited for other foliage-gleaning species, particularly in habitats and seasons in which birds feed largely on prey residing on or flying near leaf surfaces. Branch clipping is probably ill suited for species that feed heavily on either large, fast flying insects (e.g., Odonata), or on large masses of flies (e.g., Chironomidae) that seldom land on forest foliage.

We conducted roadside and trail-side point count surveys to determine whether grassland bird abundance differs along ditched and non-ditched sampling points in southwestern Saskatchewan. Savannah and Vesper Sparrows were more abundant along roads, while Baird's Sparrows, Chestnut-collared Longspurs, and Sprague's Pipits were more abundant along trails. Clay-colored Sparrows, Horned Larks, and Western Meadowlarks were detected equally along roads and trails. The lower abundance of Sprague's Pipits along roads may be attributed to the 20–30% reduction of suitable habitat associated with the road right-of-way within a point count of 100-m radius. Larger differences for Baird's Sparrows and Chestnut-collared Longspurs (42 and 56% less abundant along roads, respectively) suggest that these species tend not to establish territories adjacent to roadside ditches. Our results indicate that roadside studies designed to estimate the abundance of grassland songbirds should either include trailside counts or interpret roadside data based on the affinity of a species for roadside habitat.

We tested the null hypotheses of (1) no effect of band color and (2) no effect of number of bands worn on annual recapture rates of birds on their winter range. Results are reported from four species of granivores—Spotted Towhee (Pipilo maculatus), Golden-crowned Sparrow (Zonotrichia atricapilla), White-crowned Sparrow (Z. leucophrys), and Dark-eyed Junco (Junco hyemalis). We found no effect of the number of bands worn on return rates of these species, nor did we detect an effect of band color on return rates. In the latter case, however, the power of our tests was low, and some results hinted at possible effects of color, especially yellow bands on Golden-crowned Sparrows.

We conducted a 2-yr study of the nestling diet of Red-cockaded Woodpeckers (Picoides borealis) at three locations to determine how it varied among sites. We photographed 5939 nest visits by adult woodpeckers delivering food items for nestlings. In 1994, we located cameras near three nest cavities on the Lower Coastal Plain of South Carolina and near two cavities at the Savannah River Site, which is on the Upper Coastal Plain. In 1995, cameras were installed on the Savannah River Site and in the Piedmont National Wildlife Refuge, Georgia. The cameras recorded adults bringing 33 different types of food to nestlings. Wood roaches (Blattoidea: Blattellidae, Parcoblatta spp.) were the most common food composing 50% of the diet overall. They were also the most common prey at each location and for all but one of the woodpecker groups studied. Wood roaches were recorded in 26% of the visits photographed on the Lower Coastal Plain and 62% of the nest visits on the Upper Coastal Plain in 1994. In 1995, wood roaches were recorded in 57% and 50% of the visits on the Upper Coastal Plain and Piedmont, respectively. Woodpeckers on the Lower Coastal Plain used blueberries (Vaccinum sp.) and sawfly larvae (Hymenoptera: Diprionidae, Neodiprion sp.), two dietary items not commonly used at the other locations. Adults at two locations providing snail shells to nestlings possibly as an additional source of calcium. Morista's index of diet overlap (C) ranged from 0.94 to 0.99 for breeding males and females in the same group, from 0.63 to 0.99 among groups at the same location, and from 0.68 to 0.96 among locations. Because diet overlap of Red-cockaded Woodpecker nestlings at different geographical locations was within the range that occurred among groups at the same location, we conclude that nestling diets are similar across the geographical area studied, and that it varies little from year to year.

We modified a figure-8 leg-loop harness designed for small passerines to attach successfully 1.8-g radio transmitters over the synsacrum of migrant Soras (Porzana carolina). Because of the short caudal region of Soras, addition of a waist loop was critical to securing the transmitter while leg loops were maintained to center the package. Thin gauge (0.6-mm diameter) elastic thread proved ideal for transmitter attachment and allowed for freedom of movement and girth expansion associated with fattening during a 6–10-wk stopover. Of 110 Soras radio tagged during three field seasons, only a single mortality was observed and only a single bird lost its transmitter. Migration from the study area was confirmed for 76 (69%) and suspected for another 25 birds (total 92%).

Northern Spotted Owl (Strix occidentalis caurina) reproduction and turnover (when an owl died or shifted territories, and was replaced by another owl) were monitored at 51 locations on Simpson Timber Company lands, northwestern California, from 1991–1995. We tested for differences in proportions of five stand age classes and reproductive success between Spotted Owl pair sites with (≥1 turnover) and without turnovers. Owl pairs at sites without turnovers fledged more young, showed more consistent reproductive success, and were surrounded by a greater percentage of 21–40-yr-old stands than were owl pairs at sites with turnovers. We hypothesize that pairs with high mate fidelity and survival were more reproductively successful because those pairs had previous breeding experience together. By investigating turnover along with habitat features, we now have an indication of the relative quality of various habitats for Spotted Owls on managed, coastal forests of northern California.

Managing for the endangered Red-cockaded Woodpecker (Picoides borealis) on federal lands requires burning large tracts of mature pine forests every 3–5 yr. Many cavity trees that serve as potential nest sites for primary and secondary hole-nesting birds are destroyed by fire. We assessed the efficacy of a nest box program for the Great Crested Flycatcher (Myiarchus crinitus) at Carolina Sandhills National Wildlife Refuge, an area intensively managed for Red-cockaded Woodpeckers. During 1996–1998, we installed and monitored 330 (30 in each of 11 sites) nest boxes in mature (>60 yr) longleaf pine (Pinus palustris) tracts that were burned either in April–June (warm season) or December–March (cool season). Prescribed-burned sites were nearly devoid of snags; we estimated only 0.8/ha in cool-season burns and 1.7/ha in warm-season burns. Great Crested Flycatchers built nests in 20% of the boxes available to them. Clutch sizes were larger in warm-season burns than in cool-season burns, but fledging success (fledglings/nest hatching ≥1 egg) was lower. Twenty-two of 59 Great Crested Flycatcher nests were depredated and the proportions in each burn class were similar. We recommend the installation of nest boxes for Great Crested Flycatchers in prescribed-burned pine forests, but additional research is needed in these habitats on nest depredation rates and causes.

We modified conventional, funnel-entrance dabbling duck bait traps to increase captures for banding of American Black Ducks (Anas rubripes) in tidal saltmarsh habitats of Smith Island, Maryland, one of the few remaining strongholds for breeding Black Ducks in the Chesapeake Bay. Traps and trapping techniques were adapted to tidal creeks and refined to improve capture rate, reduce mortality, and minimize interference by gulls. Best results were achieved by synchronizing trapping with predawn, low-tide foraging patterns of Black Ducks. Trap entrances were critical to retaining ducks, and use of loafing platforms reduced overall mortality to 3% of captures per year. We captured 3071 Black Ducks during the 14-year period, 1984–1997.

The nonindigenous Eurasian Collared-Dove (Streptopelia decaocto) most likely invaded southeastern Florida from the Bahamas in the early 1980s and had established a breeding population in Dade County by 1982. In the decade after its arrival in Florida, the Eurasian Collared-Dove dispersed northward, initially along the Atlantic and Gulf coasts, and established local populations throughout the state. By 1996, the species was recorded at 46 (82%) of 56 localities surveyed on the Christmas Bird Count in Florida. The strong dispersal capability of the Eurasian Collared-Dove, which was documented during its colonization of Europe in the mid-1900s, suggests that rapid colonization of North America by the species is highly probable. Populations established in Texas and North Carolina during the early 1990s may be products of dispersal of doves from Florida. The species presents two potential biological threats to native avifauna: competition for resources and transmission of diseases.

While testing a commercially manufactured wildlife video system, we developed a method for using this technology to monitor continuously the reproductive behaviors of wild birds during the nesting period. Two passerine bird nests, Northern Cardinal (Cardinalis cardinalis) and Red-winged Blackbird (Agelaius phoeniceus), were monitored using miniature video cameras at the nests. Cameras were connected to a 24-h time-lapse video recorder located up to 18 m from each nest. System installation required approximately 20 min and nests were monitored for 20 and 15 d, respectively. Significant events recorded on videotape included: egg laying, incubation, egg hatching, nestling rearing, fledging, mammalian depredation of inviable eggs after fledging, and reptilian depredation of nestlings prior to fledging. This technology reliably provided a highly detailed, non-intrusive means of observing nesting behavior and predation events equally well during day and night operations and under extreme environmental conditions. Each commercially manufactured video system and associated equipment used in this study cost approximately US $4000.