Hybridization and introgression can promote adaptive potential and evolutionary resilience in response to increased pressures of climate change; they can also disrupt local adaptation and lead to outbreeding depression. We investigated female fitness consequences of hybridization in two sister species that are endemic to a threatened tidal marsh ecosystem: Saltmarsh (Ammospiza caudacutus) and Nelson’s Sparrows (A. nelsoni). We found increasing nest flooding rates due to rising sea levels are outpacing potential adaptive benefits of hybridization due to very low overall nesting success in both the Nelson’s and Saltmarsh Sparrows. In the center of the hybrid zone across two years, we determined the success of 201 nests of 104 pure and admixed Saltmarsh and Nelson’s Sparrow females, genotyped using a panel of Single Nucleotide Polymorphisms (SNPs) from double digest restriction-site associated DNA (ddRAD) Sequencing. We evaluated five metrics of female fitness and modeled nesting success in relation to genotypic, environmental, and nesting characteristics. We found differential fitness among Saltmarsh, Nelson’s, and hybrid females, such that birds with predominantly Saltmarsh Sparrow alleles had higher reproductive success than birds with predominantly Nelson’s Sparrow alleles, and hybrids were intermediate. Fledging success increased with two known tidal marsh nesting adaptations: nest height and nesting synchrony with tidal cycles. We found a positive relationship between hybrid index and fitness in daily nest survival in 2016, but not in 2017, likely due to differing levels of precipitation and nest flooding between years. The strongest and most consistent predictors of daily nest survival were nesting synchrony with lunar tidal flooding cycles and daily maximum tide height. Fitness patterns suggest there may be an adaptive benefit of interspecific geneflow for the Nelson’s Sparrow at the detriment of the Saltmarsh Sparrow; however, flooding rates are so high in many years they mask any fitness differences between the species, and all females had poor nesting success, regardless of genetic makeup.

We captured, banded, and genotyped a total of 104 female sparrows across the two sites and years. We monitored 202 nests (including 301 nestlings) of pure and admixed Saltmarsh and Nelson’s Sparrows across the two sites across the 2016 and 2017 breeding seasons. We conducted nest monitoring at both sites during May-August, encompassing three nesting cycles each in 2016 and 2017, as described in Maxwell et al. 2021 and following protocols from Saltmarsh Habitat and Avian Research Program (SHARP, www.tidalmarshbirds.org). Nests were visited every 3–4 days until completed and assigned an overall fate (fledged, failed due to flooding, failed due to predation, or failed due to unknown causes), following established standardized protocols (Ruskin et al., 2017). We calculated date of nest initiation based on known duration of egg-laying (3–5 days), incubation (11–12 days), and chick development (8–11 days) to determine first egg date following methods developed by Shriver et al. (2007).

We collected vegetation and nest characteristic data to test predictions about nesting characteristics as drivers of reproductive success. Vegetation data were collected within 1 m² surrounding each nest upon its completion (fledge/fail/unknown failure). Measurements included thatch depth (height of vegetative thatch layer from marsh surface in cm) at nest and across four cardinal points in 1 square meter surrounding the nest, average vegetation height directly above and surrounding the nest (in cm across four cardinal points in 1 square meter and at the center), as well as species composition of the vegetation to determine the percent high marsh vegetation surrounding the nest (percent Spartina patens). We recorded physical characteristics of the nests including, height above the ground (cm from cup lip and cup bottom to surface of the marsh), presence/absence of nest canopy (woven/domed structure that effectively covers the nest cup and prevents eggs from washing out of the nest during high tides (Humphreys et al. 2007)), percent of nest visible from above, and the species of vegetation of which the nest was made. To determine timing of nest initiation in relation to the nearest flood tide, we calculated the number of days after the new moon (the highest tidal amplitudes and flooding were on new moon dates due to lunar tidal cycles) that the nest was initiated, with initiation determined as described above, following Shriver et al. (2007).

HOBO water level loggers (ONSET, Bourne, MA) were placed at the bottom of a central channel at each study site to monitor the water levels on each day of the breeding season. These loggers measure the total pressure above their location at 15-minute intervals. With barometric pressure collected from the National Oceanic and Atmospheric Administration Stations nearest the study site locations, a compensation was made using HOBOware Pro software to determine water level seen at each marsh in 15-minute intervals throughout the entire three-month breeding season (in 2016 and 2017).

Nestlings were banded with a USGS aluminum leg band and a single site-specific color band when they were ~6 days old. Additionally, standard morphological measurements were taken on each nestling during banding, including weight, tarsus length, bill length, head length, and wing cord.

Proxys of fitness used in this study were short-term reproductive metrics taken across the two-year study period and included: daily nest survival estimates, fledging success (number of offspring fledged from each nest, including 0), hatching success (number of eggs hatched from each nest, including 0), clutch size (maximum number of eggs/nestlings in a nest), average nestling size, and maximum nestling size in a nest (measured in grams at ~ 6 days of age at banding). These measures of reproductive success represent a snapshot in time (one breeding cycle), and do not reflect lifetime fitness of an individual.

Data are provided in their raw form in columns in Excel spreadsheets. 

Data files are provided as CSV spreadsheet files and do not require any special software to open or use. From here, data can be used for modeling or for creation of capture histories for survival analysis.