Mate desertion affects offspring survival, development and physiology in a songbird with multiple parental strategies

Sexual conflict affects the amount and duration of offspring care each parent invests, resulting in multiple parental care strategies sometimes coexisting within a single population. Understanding the persistence of multiple parental care strategies requires a precise estimate of the benefits and costs associated with parental decisions. Even though the benefits of brood desertion are well known, the reproductive costs of desertion (i.e., nestlings’ physiological conditions and survival) are less explored. We use rock sparrows, Petronia petronia, a species in which both uniparental and biparental care occur in the same population, to investigate the costs of brood desertion. Specifically, we continuously monitored breeding attempts to explore the behavioral mechanisms (desertion decision and compensatory responses) and the reproductive and physiological consequences (offspring corticosterone concentrations, oxidative stress, telomere attrition; mechanistic) of parental care strategies. We show that male desertion was not related to the initial value of the brood (clutch size, brood size) but was associated with a reduction in the survival probability of the nestlings. Females caring alone increased their per capita feeding rate, partially compensating for the lack of male care. Nestlings deserted earlier also experienced higher oxidative stress and had higher corticosterone concentrations during the early stages of development, but these effects did not persist to fledging, and there were no differences in telomere attrition. Our findings indicate combined reproductive and physiological costs associated with brood desertion. Considering these costs is essential to understanding the evolution and persistence of polymorphic patterns of care.

Methods
Experimental Design:

We conducted our study from May to July 2021 and 2022 on a wild rock sparrow population breeding in nest boxes on the grassland slopes of Valsaín (Segovia, central Spain, 40°53′74″N, 4°01′W, 1200 m above sea level). This population has been monitored since 2013 (Cantarero et al. 2019; Corregidor-Castro et al. 2022). We routinely checked nest boxes for clutch initiation and determined the onset of egg laying, incubation, and hatching. After hatching (day 1), we recorded videos of all nests to quantify provisioning and desertion behavior on days 3, 6, 8, 11, and 14, for 1.5 hours each (Figure S1). Digital video cameras (Sony Handcam CX405) were placed approximately 20 m away from the nest boxes such that they would capture a ~ 2 m radius surrounding the nest box. To minimize observer bias, blinded methods were used when all behavioral data were recorded and analyzed. During each filming session, we recorded hourly provisioning rates (visits to the nest per hour) by the male and female, individually. We determined a nest to be deserted by the male when the male was not present during the entire duration of the video (Griggio, Matessi & Pilastro 2005), and did not reappear in any subsequent videos throughout the development period. The time of desertion was estimated as the midpoint between the two consecutive recordings in which a male was present and then not present (Griggio, Matessi & Pilastro 2005). During the two years of study, only male desertion was observed (no occurrences of female desertion). All uniparental nests in the present study, therefore, represent female-only care.

Corticosterone:

We measured offspring plasma corticosterone level using enzyme-linked immunosorbent assays (Enzo Life Sciences; Farmingdale, NY, USA) following manufacturer instructions, and read optical density on a plate reader at 405 nm. We first validated this assay for rock sparrows by using serial dilutions of plasma with two different concentrations of steroid displacement reagent (SDR; 0.5% and 1% of plasma volume) and a standard curve. From this, we determined an optimal plasma dilution of 1:10 with 0.5% concentrated SDR for subsequent assays. The Samples order was randomized across plates. Intra- and inter-plate CV, calculated from plasma run in triplicate, were 17.6% and 16.8%, respectively.

Oxidative Stress:

To measure reactive oxygen metabolites (dROMs), we used a kit for detecting hydroperoxides (MC003, Diacron, Grosseto, Italy), which signal both protein and lipid oxidative damage (Consantini and Dell’Omo, 2006). Briefly, we diluted 2µl of blood plasma in the provided acidic buffered solution (1:50) and followed the end-point mode protocol from the manufacturer (with modifications for use on 96-well microplate) and read optical density on a plate reader at 546nm. Samples were run in duplicate, and intra- and inter-assay variation was 4.8% and 15.9%, respectively. To measure antioxidant capacity (oxy), we used an OXY-adsorbent test (MC 435; Diacron, Grosseto, Italy), which measures blood antioxidant barrier by quantifying the oxidant action of hypochlorous acid (HClO; Vassalle et, 2008). Briefly, we diluted 2µl of blood plasma in distilled water (1:100) and followed the manufacturer's protocol (with modifications for use on a 96-well microplate). We then mixed a 5µl subsample of diluted plasma with 195µl of proved HClO solution. All samples were run in triplicate, with calibrators distributed vertically and horizontally across the plate. Intra- and inter-assay variation was 8.1%% and 5.3%, respectively.

A biologically relevant assessment of oxidative stress includes both the levels of circulating dROMs and the absorbance capacity of the blood (Oxy) within individuals. So, we used an integrative index of oxidative stress to combine these measures (Vassalle et al. 2008). We standardized dROM and oxy using the function ‘scale’ and then calculated the difference between scaled values for each individual, such that higher values of the oxidative stress index correspond to a greater differential between dROM and oxy.

Telomeres:

Genomic DNA was extracted from erythrocytes using the Qiagen DNeasy Blood and Tissue Kit, following the manufacturer's protocol with a modification: the wash with buffer 2 was repeated twice. Sample elution from the column was also performed twice with 30 µl of AE buffer provided in the kit. DNA quality of each elution was assessed using a Nanodrop ND-2000 C spectrophotometer (Thermo Scientific, USA), with 260/280 and 260/230 ratios greater than 1.8 considered acceptable. DNA quantification was performed using Qubit (Invitrogen) with the AccuGreen Broad Range dsDNA quantification kit (Biotium). Samples were diluted to a concentration of 3 ng/µl.  Relative telomere length (RTL) was measured using quantitative real-time PCR (qPCR) (Cawthon, 2002), calculating the ratio between telomeric DNA and the single-copy reference gene Glyceraldehyde-3-phosphate dehydrogenase (GAPDH). Telomeric primers Tel1b and Tel2b, and GAPDH primers GAPDH-F and R were employed (Criscuolo et al., 2009). Amplifications were performed on a BioRad CFX384 Touch Real-Time PCR System. Each reaction was conducted in a 10 µl volume, containing 2 µl of 5× HOT FIREPol® EvaGreen® qPCR Mix Plus (without ROX) (Solis BioDyne), 6 ng of genomic DNA, and 200 nM of each forward and reverse primer. The qPCR profile consisted of an initial step at 95°C for 12 minutes, followed by 40 cycles of 95°C for 20 seconds, 58°C for 18 seconds, and 72°C for 1 minute. Upon completion of each run, a melt curve (65 to 95°C, with 0.5°C increments for 5 seconds) was generated to confirm qPCR specificity. Each plate contained samples collected from nestlings at 8 and 14 days after hatching, two interplate calibrators, and a negative control for both telomere and GAPDH, all run in triplicate. Baseline and cycle quantification (Cq) values were corrected using LinRegPCR software ver. 2017.1 (Ruijter et al., 2009). Between-run variation was removed using Factor qPCR (Ruijter et al., 2015). Relative telomere length was calculated following the equation proposed by Pfaffl (Pfaffl, 2001), as reported in Morbiato et al. (2023) and in Monteforte et al. (2020). An acceptance threshold for amplification efficiency was set at 100 ± 20%. Interassay coefficients of variation (CV) were 3.7% for telomere and 1.3% for GAPDH, while intraassay CVs were 1.3% for telomere and 1.4% for GAPDH.