Since the early 1990s, aerial insectivorous birds have shown serious population declines in North America, but it is not clear if factors common to all species within this guild account for these declines. Among sympatric swallows, population trends differ, and this may be due to differences in ecology operating throughout the annual cycle. Although these species all feed on aerial insects, prey taxa can differ tremendously in their “aeroecology” and use by swallows. We examined the potential for dietary differences among three species of swallows, Barn Swallow (Hirundo rustica), Cliff Swallow (Petrochelidon pyrrhonota), and Tree Swallow (Tachycineta bicolor), breeding sympatrically in southern Ontario, Canada. Potential interspecific differences in nestling diet were examined using two endogenous biomarkers, DNA barcoding of nestling feces and stable isotope analysis (δ 2H, δ 13C, δ 15N) of nestling feathers. We found evidence for differences in dietary sources of provisioned young where Barn Swallows provisioned more terrestrial-based prey, Cliff Swallows provisioned an intermediate diet, and Tree Swallows the most aquatic-emergent insect diet. We suggest this information may help to identify potential factors contributing to differential declines of aerial insectivores operating on the breeding grounds, including diet quality.

Fieldwork was conducted at nine farms from May-July 2018 within Wellington County, near Guelph, Ontario, Canada (43.55° N, 80.25° W). This area is a rural landscape characterized by a mixture of row crops, pasture, small woodlots, and a few large conservation areas (Kusack et al. 2020). The region includes one large lake (Guelph Lake) and numerous small wetlands and ponds (Kusack et al., unpublished data). These locations were grouped into seven sites because two sets of two farms were adjacent to each other such that we considered them the same site. Six of these sites had all three species of swallow breeding sympatrically (average 2018 clutch initiation was 25 May for Barn Swallows, 29 May for Cliff Swallows, and 24 May for Tree Swallows). However, Cliff Swallow nests failed at one of the sites, so only Barn and Tree Swallow samples were taken there. At another site (Guelph Lake Conservation Area), only Tree Swallows were present. All Barn Swallow and Cliff Swallow colonies were located inside barns, whereas Tree Swallows were in nest boxes.

Nest Monitoring and Sampling

Nest monitoring began in early May and involved visiting sites one to two times weekly to establish the approximate onset of laying. If eggs were present upon first inspection, and the clutch was not complete, it was assumed that one egg had been laid per day for all three species (Winkler et al. 2011; Brown et al. 2017; Brown and Brown 2019). Once the clutch initiation date was estimated, nests were monitored at least twice per week to determine final clutch size and approximate hatch date. Timing of hatching was predicted based on a 14-day incubation period after the penultimate egg was laid for Barn Swallows (Brown and Brown 2019), Cliff Swallows (Brown et al. 2017), and Tree Swallows (Winkler et al. 2011). To determine hatch day as accurately as possible, nests were visited before predicted hatch day and every few days thereafter. Hatch Day was assigned based on evidence of hatching; this included eggshells present, unhatched eggs remaining, and nestlings still being wet. Hatch Day was considered Day Zero. For nests where hatch day could not be assigned, nestlings were aged by feather tract development (Stoner 1935; Stoner 1945; Marsh 1980). We only focused on sampling first clutches.

All nests were visited between Day Six and Ten after hatching, during which Tree Swallow and Barn Swallow nestlings were banded with a uniquely numbered United States Geological Survey (USGS) aluminum leg band. For each nestling, mass, age and relaxed wing chord were documented. Fecal samples were collected by holding a clean piece of paper under the young as they were removed from their nest, these samples were grouped by nest. Cliff Swallow nestlings were not accessible for banding and therefore cardboard was placed beneath the nest for roughly 30 minutes to collect fecal samples while we processed other nestlings from the same site.

Barn Swallow and Tree Swallow nests were visited and morphometric measurements taken when young were 15 days old. It was not possible to visit all nests at Day 15 and so some Tree Swallows were measured at Day 16 or 17. At that time, feathers were taken from all but one nestling per nest for stable isotope analyses. Hatch-year Cliff Swallows were visited close to fledging (~Day 20) and when they could be coaxed from the nest structure or accessed by shortening the entrance tube. At that stage, they were also weighed, measured, and had a central tail feather taken for stable isotope analyses, then returned to their nests. After sampling, young were expected to fledge within 1 - 7 days. Average fledging age is 19-20 days, 20-21 days, and 18-22 days for Barn Swallows (Brown and Brown 2019), Cliff Swallows (Brown et al. 2017), and Tree Swallows (Winkler et al. 2011), respectively. All young were banded with a USGS numbered band during feather sampling.

Stable Isotope Analyses

Tail feathers were soaked with 2:1 chloroform:methanol overnight, rinsed and dried in a fume hood at ambient temperature for 24 h. Samples analyzed for stable hydrogen (δ2H) isotopes were weighed (0.35 mg) into silver capsules using the feather barbs only. Capsules were compressed and analyzed using the LSIS-AFAR stable isotope facility at the University of Western Ontario. Samples were loaded into a Uni-prep carousel (Eurovector®, Milan, ITA) held at 60ºC, evacuated and maintained under pressure with dry helium and then combusted in a Eurovector 3000 elemental analyzer (Eurovector, Milan) pyrolytically on glassy carbon at 1350ºC. Separated H2 was analyzed using a Thermo Delta V Plus (Thermo scientific®, Bremen, DEU) continuous-flow isotope ratio mass spectrometer via a Conflo device (Thermo Scientific, Bremen, DEU). Sample results were expressed in the standard delta (δ) notation in parts per thousand (‰) deviation from the Vienna Standard Mean Ocean Water (VSMOW) standard. In-house keratin standards (CBS: -197‰; KHS: -54.1‰) were used to derive the δ2H value of the non-exchangeable H fraction according to the comparative equilibration approach (Wassenaar and Hobson 2003). Based on within-run (n = 5 each) keratin standards, measurement error was estimated to be ±2‰.

Feather samples analyzed for stable carbon (δ13C) and nitrogen (δ15N) isotope values were weighed (1.0 mg) in tin capsules, compressed and analyzed at the Environment and Climate Change Canada stable isotope laboratory in Saskatoon, Saskatchewan, Canada. Samples were combusted at 1030°C in a Carlo Erba NA1500 (Thermo Scientific; Waltham, United States) or Eurovector 3000 (Eurovector, Milan) elemental analyzer. The resulting N2 and CO2 were separated chromatographically and introduced into an Elementar Isoprime (Elementar; Langenselbold, Germany) or a Nu Instruments Horizon (Nu Instruments Ltd.; Wrexham, United Kingdom) isotope ratio mass spectrometer. Sample results were expressed in the standard delta (δ) notation in parts per thousand (‰) deviation from international standards (Vienna Pee Dee Belemnite [VPDB] and AIR for δ13C and δ15N, respectively). Internal laboratory calibration standards were BWBIII keratin (δ13C = -20.18‰, δ15N = 14.31 ‰) and Pugel (δ13C = -13.64‰, δ15N = 5.07‰). Measurement precision was based on replicate (n = 5) within-run measurements of internal reference material and estimated to be ±0.1‰ for both δ13C and δ15N.

Fecal DNA Analyses

Fecal samples were transferred in the field into scintillation vials containing 95% ethanol and kept cool with ice packs. Upon returning from the field, samples were immediately frozen at -20^oC and later processed at the Canadian Centre for DNA Barcoding (Guelph, Canada). Samples were amplified separately, using arthropod-specific primers targeting a 157 base-pair region of the mitochondrial cytochrome c oxidase subunit 1 (COI) gene (Hebert et al. 2003). Amplified samples were pooled and sequenced using an Ion Torrent PGM high-throughput sequencer (Thermo Fisher Scientific), trimmed to remove the primer, and filtered for a minimum size of 100 base pairs. Filtered reads were queried against the Barcode of Life Database (BOLD; www.boldsystems.org) reference using a basal local alignment search tool (BLAST) algorithm to assign taxonomic identity. Results were accepted if they had a minimum of 50 reads that matched reference sequence with at least 95% identity across at least 100 base pairs.

Bumelis_Fecal-DNA:

This file contains information for prey species found in nestling fecal samples. Column B shows which species was sampled (BARS = Barn Swallow, CLSW = Cliff Swallow, TRES = Tree Swallow). Columns C-E show the age (days since hatching), site code, and collection date. Column F is the number of reads for each species, and Columns G-L identify the prey to the lowest taxonomic level possible.

Bumelis_Feather-Isotopes

This file contains information on nestling feather isotope values. Column B shows which species was sampled (BARS = Barn Swallow, CLSW = Cliff Swallow, TRES = Tree Swallow). Column C&D are site and date information. Columns E-G are stable isotope values (‰)