Immune defense is fundamental to diminish the negative effects of the attack of infectious agents, yet the activation of the immune system entails costs and may compromise other life history-traits such as reproduction. In reproductive brown booby pairs (Sula leucogaster), we experimentally imposed an immune challenge during incubation, by intraperitoneally injecting Escherichia coli lipopolysaccharide (LPS), in either the male or the female. We aimed to test whether activation of the immune response results on: (1) an increase in oxidative stress parameters, (2) a decline in post-hatching parental care in the treated individual, and (3) a compensation of the post-hatching parental effort by the non-treated mate. We found that activation of the immune response during incubation did not increase oxidative damage to lipids or total antioxidant capacity. However, mounting an immune response compromised parental effort during the chick rearing period: compared to controls, LPS-treated parents showed roughly a twofold decline in the rate of offspring feeding and preening. Interestingly, mates of LPS-treated parents increased 58% their feeding rate suggesting parental care compensation. According to a scenario of full compensation, the decline in parental effort of LPS-treated parents did not result in a poorer offspring growth or immune response, or increased levels of oxidative stress parameters. These findings suggest that in a long-lived species with long-lasting biparental care, an immune challenge compromises parental care, favoring parental compensation as a strategy to mitigate costs in terms of offspring success.

Summary of Methods

The study was conducted in the brown booby breeding colony at Isla Larga, Parque Nacional Islas Marietas, Nayarit, México.

Please note that "N/A" in the database indicates a value is not available.

Experimental immune challenge

During incubation, 61 breeding pairs were captured and individually marked with a leg band. We randomly assigned pairs to the LPS group (n=41) or the control group (n=20). In the LPS group, one member of each pair (i.e. LPS-treated parent, females in 20 pairs and males in 21 pairs) was intraperitoneally injected with 0.10 mg of E. coli LPS (serotype O55:B5, Sigma Aldrich, catalogue L2637, St. Louis, MO, USA) dissolved in 1 ml of Phosphate Buffered Saline solution (PBS; Sigma Aldrich, catalogue P4417, St. Louis, MO, USA). In the control group, one member of each pair was sham-treated (i.e. control-treated parent, females in ten pairs and males in ten pairs) with an intraperitoneal injection of 1 ml of PBS. Ten days later, we applied a second LPS injection to the LPS-treated parents to activate a secondary humoral immune response and a second PBS injection to control-treated parents. LPS injections were administered on average 28 ± 12 and 18 ± 12 days before hatching for the first and the second dose, respectively. LPS is a conserved pathogen-associated molecular pattern (PAMP) with a highly immunogenic component of Gram-negative bacteria cell walls that mimics a bacterial infection and has been widely used in ecoimmunological studies.

To determine oxidative stress parameters, we collected three 1.5 ml blood samples from the brachial vein of LPS- and control-treated parents: (1) before the first immune challenge (during the first capture), (2) ten days (± 2 days) after the first challenge, before the second immune challenge (second capture), and (3) seven days (± 2 days) after the second immune challenge (third capture). Immediately after collection, blood samples were kept on ice, plasma and blood cells were separated by centrifuged at 10000 g × 10 min, and plasma was stored at -80 ºC until laboratory analyses were performed. To determine the hatching date, the nests were checked daily between 14:00 and 16:00 hrs. This time period was chosen as it is the time of less activity, to avoid disturbing the colony. Bird handling took less than 12 min to minimize stress.

Offspring growth and immune response

In the offspring, we measured the increase of body mass (± 10 g), tarsus, ulna, and beak length (± 0.1 mm) every 5 days from hatching until 30 days post-hatching. For oxidative stress analyses, we collected one 1 ml blood sample from 15-days-old offspring. As an estimate of the immune response of chicks, we used the phytohemagglutinin skin test that includes both, adaptive and innate components of the immune system. At the age of 30 days, offspring were injected subcutaneously in the left wing-web with 0.1 ml of 1 mg ml^-1 PHA-P (Sigma Aldrich, catalogue L9016, St. Louis, MO, USA) dissolved in PBS. The point of injection was marked with an indelible marker and wing-web thickness was measured (three times to yield an average) before injection and 24 hours later with a Mitutoyo micrometer (± 0.001 mm). We estimated the immune response to PHA as the wing-web swelling 24 hours after the injection. The repeatability of swelling measures was high (intraclass correlation coefficient ± SE: before injection, r =0.96 ± 0.008, P < 0.001; after injection r = 0.99 ± 0.001, P <0.001).

Behavioral observations

Observations of parental care started when the chicks were roughly 4 days old and lasted until the chicks reached an age of 30 days. Nests were observed every five days (± 3 days), from 7:00 - 9:30 hours and 18:30 - 21:00 hours, the periods of greater diurnal parental activity at the colony. Behavioral records were conducted by seven observers from roughly 3 - 6 m of distance from the focal nests. Observers were blind about which experimental treatment focal nests belonged. All nests recorded had only one chick after the second observation. Parental behaviors recorded were: (1) time that each parent spends near the chick (hereafter offspring attendance; recorded as the minute and hour a parent arrived and left the nest), (2) absolute number of feedings by each parent, (3) number of bouts of preening behavior from each parent to chick. From the focal offspring, we recorded (4) the number of begging bouts (when the chick raises the head and vocalizes with a “tac-tac” sound; a new begging bout was considered after a period of >10 seconds of no begging). We pooled all observations from each nest and calculated the time each parent spent with the offspring, and the rate per hour of each parental behavior (preening rate = absolute number of preenings by each parent / total time the nest was observed; feeding rate = absolute number of feedings by each parent / total offspring attendance by the male or the female). We independently calculated begging rates for the mother and father (number of begging bouts to the male or the female / total offspring attendance by the male or the female). We started behavioral recordings after a period of observers training, when inter-observer reliability was high (r = 0.97± 0.02, P = 0.02).