The study was performed from 2012 to 2014 in the Hoya of Guadix-Baza, Granada province, southeast of Spain (37°18′N, 3°11′W). The area is an extensive agricultural landscape with scattered holm oaks (Quercus ilex) where cork-made nest-boxes were placed to favour the reproduction of a cavity-nesting bird community (for more details see41) including scops owls.
The scops owl is a medium-sized, nocturnal and trans-Saharan migrant owl arriving from their winter African quarters to the study area in April49,55. In the area, scops owls begin reproduction throughout May55, making one clutch per year of about 2–6 eggs. Incubation takes 24–25 days, and is performed by the female. Both sexes participate in feeding tasks through the nestling period49. Plumage coloration in the species varies from grey to brown so that individuals can be assigned to one of two distinct color morphs in relation to the amount of phaeomelanin: greyish or brownish.
Field data collection
Every year, from the beginning of the breeding season, nest-boxes were visited once a week until egg-laying was detected. After detection of a breeding attempt, nests were visited once more just before the estimated hatching date to capture the incubating female by hand during incubation. After hatching, nests were monitored by weekly visits and at the end of the nesting period, we recorded the number of fledglings and fledgling weight. Males were captured at nights by means of nest-traps, when they brought food to offspring. All individuals were ringed with individually numbered metal rings and sexed based on inspection of the brood patch (only present in females49). Moreover, upon capture, all adults were photographed for color characterization as in Parejo, et al.56. In brief, by focusing on redness extension at the head, breast and wings–back, we scored each body part among 1 to 3 points depending on the degree of redness56. At the end, scores of the three body parts were summed to get an individual score for every bird ranging from 3 to 9. Individuals were then classified as either brownish (score >7) or greyish (score ≤7).
When the youngest chick of each nest was 12 days old, we extracted blood from nestlings to obtain innate (agglutination activity) immunity measurements. Scops owl chicks hatch asynchronically so that chicks from the same nest may have slightly different ages. Blood was collected from the brachial vein using a 0.5 × 16 mm needle and heparinized capillary tubes to later transfer it into a 1.5 mL tube. Samples were refrigerated immediately after extraction, and later on the same day, plasma and red blood cells were separated by centrifugation at 13300 rpm for 5 min. All samples were stored in a −20 °C freezer until the end of fieldwork when laboratory analyses were performed. Finally, fledglings were weighed to the nearest 1 g using a 300 g Pesola spring balance just before fledging. Ten days later, nests were re-visited to verify fledging success. Nestlings not found dead in the nests during that last visit were considered to have fledged.
Animal data collection complies with the current laws of Spain and the fieldwork including adult trapping and obtaining blood samples was authorized by Consejería de Medio Ambiente y Ordenación del Territorio de la Junta de Andalucía (projects CGL2011-27561/BOS and CGL2014-56769-P; licence code: P06-RNM-01862). The study protocol was reviewed and approved by the ethical committee of the Spanish Research Organism (CSIC).
Assessment of innate humoral immunity was done by using the standard protocol based on Natural Antibody (hereafter NAb) mediated complement activation and red blood cell agglutination57.The agglutination responsiveness represents the interaction between NAb and antigens. Quantification of agglutination was done by assessing the dilution stage (on a scale from 1 to 12 titres) at which this reaction stopped against the same amount of rabbit blood cell suspension on digitalized images (for more details on the method see57,58). This assay determines the values of circulating NAbs by measuring red blood cell agglutination. We did not consider lytic activity, which can be feasibly determined by this assay, because it did not show variation in nestlings.
Experimental manipulation of social information on predation risk
Nests with hatchlings were randomly assigned to one of the following treatments: (i) “Alarm” (N = 20 nests), in which we broadcasted little owls’ alarm calls through all the nestling period to simulate threatened little owls; (ii) “Non-alarm” (N = 17 nests), in which we broadcasted non-alarm calls of little owls through all the nestling period to simulate the presence of non-threatened little owls; and (iii) “Control” (N = 14 nests), in which no playback was broadcasted but visits were as frequent as to alarm and non-alarm nests. Therefore, Alarm and Non-Alarm nests are informed nests, while Control nests are uninformed ones.
Neither laying date (General lineal model with a Normal error distribution: treatment effect: F2,42 = 0.25, P = 0.78; year effect: F2,42 = 19.71, P < 0.0001; treatment * year effect: F4,42 = 1.35, P = 0.27), nor clutch size (General linear model with a Poisson error distribution: treatment effect: F2,42 = 0.05, P = 0.95, year effect: F2,42 = 0.45, P = 0.64; treatment * year effect: F4,42 = 0.06, P = 0.99), nor female morph (Logistic regression model: treatment effect: χ2 = 1.54, P = 0.46; year effect: χ2 = 2.02, P = 0.36; treatment * year effect: χ4 = 4.82, P = 0.31) differed among nests assigned to the different treatments, indicating that the experiment was correctly randomized.
In the second and third years of the study, all territories that were reused by scops owls (5 territories) were assigned to one alternative treatment of the previous year to discard the possibility of biased results due to differences in territorial quality. Furthermore, none of the reused territories were consecutively occupied by the same individuals; therefore, a possible effect of the reutilisation of territories by the same individuals in our experiment can be discarded. On the other hand, the 10 females that were breeding in the population in consecutive years were assigned to alternative treatments in the different years to avoid accumulative effects of the experiment. Nevertheless, we acknowledge that the inclusion of some females more than once in the experiment could affect results due to the non-independence of observations of the same female and to carry-over effects. Therefore, we analysed the effects of the treatment on parental fitness twice, first including data from all the recorded reproductive events and, second including females only the first year they were involved in the experiment. Results were qualitatively identical, and hence we reported in the results section those performed on all the reproductive events (including non-repeated and repeated females) and in the Supplementary Material (Table S2) those performed on the limited data set (including non-repeated females).
Call treatments were applied five times to every nest each 3 days from the day all eggs had hatched. Calls were broadcasted close to scops owls’ nests during 2 h each day after dusk. For that purpose, we used portable amplified speakers connected to digital takeMS audio players (model deseo) as in5,41. Three non-alarm and three alarm calls from different individuals were used to generate two distinct 1.5–3 min of uncompressed audio files with the recordings of the various calls. These two audio files were randomly mixed with periods of silence from 3 to 8 min and then randomly broadcasted to reduce pseudoreplication. Little owl calls and silent periods were recorded in separate tracks so that the exact sequence of calls and silences was randomly decided by selecting the random playback mode. The randomized presentation of the three calling bouts of each type in combination with the silence tracks during the 2 h of broadcasting produces an unique assortment of calls for each day of treatment and nest (see for similar approaches5,59,60). Average broadcasting volume was 89.1 (+1.1) dB measured 1 m away from the speaker, which closely resembles by ear to the natural production of real little owl calls.
In addition, in 2014, coinciding with the second visit to nests for call broadcasting, we recorded parental behaviour by filming nests within the same day before (during 60–90 min just after sunset, which is just after dusk) and after the beginning of the broadcasting (during 60–90 min once the broadcasting had begun). Thus, each nest was the same day first filmed in control conditions (pre-treatment time) and afterwards under experimental conditions (during-treatment time). Therefore, a potential effect of time of day on foraging remains to be controlled for. Infrared mini cameras (KPC- S500, black and white CCD camera, Esentia Systems Inc.) (Size of 25 mm (W) × 25 mm (H)) were used to record parental activity at night. Cameras were installed in the middle of the roofs of unused nestboxes that were interchanged with the roofs of the target nestboxes the filming day prior to monitoring, so that they were not likely to be detected by birds due to their small size. All the other necessary material for recording such as cables, batteries and recording discs and more likely to be detected by birds was camouflaged in the tree supporting the nestbox. Females had been marked in the head using white Tippex before filming, which allowed us to identify parents in video recordings. One observer, who was blind to time (i.e. before or during treatment) and to the treatments assigned to nests (Alarm N = 7 nests, Non-alarm N = 6 nests and Control N = 5 nests), extracted from recordings: (i) Latency as the time elapsed from the onset of filming until each of the parents return to normal activities such as brooding, cleaning or feeding at nests. In the case of males, this latency usually reflects the time elapsed until they resume bringing prey to the nest. However, at the time we recorded parental behaviour, females spend a lot of time inside nest-boxes brooding chicks and they only sometimes leave the nests to hunt. While females are in nests they usually brood nestlings, get and distribute prey brought by males among nestlings. Therefore, we measured latency for females as the time they take on their own to return to normal activities at nests, i.e. brooding, cleaning or feeding. (ii) Parental provisioning rate as the number of prey per hour of the adults separately; and (iii) mean prey size per observation period. The size of prey was estimated by comparing their length to the length of adult’s head, i.e. giving a value of 1 to prey as long as the head of the adult and relativizing the others. The difference between during- and pre-treatment periods in all these variables for a given nest were used as dependent variables in all analyses.
Analyses were performed using SAS v.9.4 statistical software (SAS 2002–2008 Institute, Cary, NC, USA).
We first analysed whether laying date (by performing a General lineal model, GLM procedure in SAS with normal error distribution and identity link function), clutch size (by performing a General lineal model, GLIMMIX procedure in SAS with Poisson error distribution and log link function) and female color (by performing a logistic regression model, GENMOD procedure in SAS) differed among nests assigned to the different treatments, and hence whether the experiment was correctly randomized. In all these models, to take into account environmental effects, the year and the interaction between year and treatment were introduced as fixed factors
Then, we performed General lineal models to test for the effect of the experimental treatment on the number of fledglings per nest (Genmod procedure in SAS with Poisson error distribution, log link function), mean weight of fledglings per nest (GLM procedure in SAS with Normal error distribution, identity link function) and mean NAb levels (GLM procedure in SAS with Normal error distribution, identity link function) per nest. In these models we introduced the treatment (with three levels: alarm, non-alarm and control) and the female color (with two levels: brownish versus greyish) as factors, and the interaction treatment × female color to evaluate the expected possibility of changing responses to the experiment with female color. In addition, laying date was introduced in models as a covariate to account for individual quality and the year as a fixed effect to take into account environmental effects. When studying the number of fledglings per nest, only nests that avoided predation were included.
We performed General lineal models (GLM procedure in SAS, Normal distribution, link = identity) with the variables measuring parental care (change in latency, provisioning rate and mean size of provisioned prey from pre-treatment to during-treatment time) as dependent variables in each model, and the treatment, individual color and its interaction as explanatory variables. In all models, laying date was introduced as a covariate to account for individual quality.
Pairwise differences in significant models were checked by comparisons of least-squared means of each treatment.